289 105 1MB
ENGLISH Pages [253]
Tartu Semiotics Library 8
Tartu Semiootika Raamatukogu 8 Тартуская библиотека семиотики 8
Geenid, informatsioon ja semioos Charbel Niño El-Hani João Queiroz Claus Emmeche
Гены, информация и семиозис
University of Tartu
Genes, Information, and Semiosis
Charbel Niño El-Hani João Queiroz Claus Emmeche
Tartu 2008
Book series Tartu Semiotics Library editors: Peeter Torop, Kalevi Kull, Silvi Salupere Address of the editorial office: Department of Semiotics University of Tartu Tiigi St. 78 Tartu 50410, Estonia http://www.ut.ee/SOSE/tsl.html
Cover design:
Copyright: University of Tartu, 2008 ISSN 1406–4278 ISBN Tartu University Press www.tyk.ee
Preface Looking at life from the perspective of the natural sciences, one often focus on life’s material, that is, the molecular composition as well as all the physical and chemical processes involved in cellular metabolism and physiological regulation of organic growth and self-maintenance processes. For a newcomer to the problems discussed in this book, it may seem strange that contemporary biology — given its success in describing heredity and evolution, its deep understanding of functional mechanisms and the composition of cells, organelles and molecules, and its present aim of synthesizing those findings into a coherent ‘systems biology’ — could gain anything from a field usually considered to be ‘humanistic’, namely semiotics, the study of sign systems, often understood as merely human constructs. Yet, traces of a human origin of some kinds of signs used in science and scientific investigation (such as indices of a measuring instrument or the symbols of a theory), though obvious, are far from sufficient to explain the origins of the objects and processes that scientists within physics, biology and their sub-disciplines are committed to investigate. The objects of biology, the subject matter of the biosciences, are not adequately described simply as a matter of complex configurations of physical or biochemical processes. This is at least a major tenet of biosemiotics, an attempt to take biological sciences, including systems biology, a step further by expanding its theoretical domain by means of a semiotic (sign-theoretic) understanding of biological processes at many levels of complexity. The gene is not simply the same as DNA understood as a chemical substance. The gene is a much more complicated notion, as we shall see. We contend that the gene will not be fully understood in biology at the deepest theoretical level unless a profound grasp of semiotics is applied to analyze genes in their function as complex signs for the production of proteins. That basic and controversial argument is not our invention, but is something we see as part of a growing cross-disciplinary science of biosemiotics. What we have done here is to state it as precise as possible, to bring it out coherently, against a long and perplexing
6
Genes, Information, and Semiosis
history of debates (within biology, genetics, molecular biology, biochemistry, and the scholarly field of history and philosophy of biology) concerning the status of the gene, or the genetic information systems, as we prefer to say. By combining semiotics and biology in a stepwise fashion, we have tried to make it possible for newcomers from both areas to follow our basic argument in detail. Through a theoretical analyses of informational processes in this part of biology, integrating central insights from general semiotics and molecular biology, we construct a detailed semiotic model of the genetic information system. Assessing the fruitfulness of biosemiotics in general will crucially depend on that field’s ability to produce models of specific biological systems that contribute to deepen our understanding of those systems’ biology and biocomplexity in general, and pose new solvable problems for further scientific research. We could not have made this contribution to flesh out in detail the fundamental structure of the argument, had we not been in continuous contact with many excellent colleagues working in this field and its neighbourhoods, and we deeply appreciate the stimulating feedback and thoughtful comments we have received during this process. That we discovered new and, we think, important lessons also for semioticians and information theorists as a spin-off of this inquiry came only as an exciting surprise to us. This book initially grew out of a six-month period Charbel Niño El-Hani spent as a researcher in the Center for the Philosophy of Nature and Science Studies, University of Copenhagen, with Claus Emmeche, with the support of the Brazilian National Research Council (CNPq). From September 2003 to March 2004, we worked out the first version of the arguments in this book, which were presented at the Danish Philosophy Society Annual Meeting, in February 2004. At this time, João Queiroz was already connected with the Research Group in History, Philosophy, and Biology Teaching, which Charbel coordinates at the Institute of Biology, Federal University of Bahia. João’s collaboration in this project began while Charbel still was in Denmark, that is, at the very first steps of the work. Back in Brazil, Charbel and João began a closer collaboration at the Federal University of Bahia. Subsequent versions of the arguments were elaborated, then, by the three of us, resulting in presentations in several meeting, such as VI Brazilian International Meeting on Cognitive Sciences (2004), in São Luís, Brazil; IV Gatherings in
Preface
7
Biosemiotics (2004), in Prague, Czech Republic; V Gatherings in Biosemiotics (2005), in Urbino, Italy; III Conference on the Foundations of Information Science (2005), in Paris, France, and VIII Gatherings in Biosemiotics, in Syros, Greece (2008). We are indebted to the audiences of these several presentations for their fruitful comments. In 2005, the journal S. E. E. D. published a first paper reporting parts of our results. A second paper, with a more developed version of the arguments then appeared in Semiotica, in 2006. This paper received the Mouton d’Or Prize, given to the best paper of the year published in the journal. Finally, a yet more developed set of results is now presented here. We are thankful, in particular, to Kalevi Kull for his efforts towards its publication. Here, you will also find some results from other papers we published, particularly in Transactions of the Charles Sanders Peirce Society, Acta Biotheoretica, Cognitio, S.E.E.D. (Semiotics, Evolution, Energy, and Development), The American Journal of Semiotics. We would also like to thank the Brazilian National Research Council (CNPq), The State of São Paulo Research Foundation (FAPESP), The State of Bahia Research Foundation (FAPESB), and the Faculty of Science, University of Copenhagen for the support given to the development of this project.
Table of contents
9
Table of contents Chapter 1. ‘Genes’ and ‘information’ as conceptual problems ................................................................................... 13 Chapter 2. The problem of the gene ...................................... 1. The birth of the ‘gene’ as an instrumental concept .............. 2. The fruitful tension between instrumentalist and realist views of the gene .................................................................. 3. Genes as units in Mendelian genetics ................................... 4. The classical molecular gene concept .................................. 5. The golden years of molecular genetics ............................... 6. How the gene concept became a problem: why is the gene not a structural unit? ............................................................. 7. How the gene concept became a problem: why is the gene not a functional unit? ............................................................ 8. What is a gene, after all? ...................................................... 9. The gene in the era of genomics and systems biology ......... 10.Some reactions to the problem of the gene .......................... Conceptual variation and ambiguities in the gene concept .. Searching for a single, inclusive gene concept ..................... The process molecular gene concept .................................... Gene-P and gene-D .............................................................. Genes as sets of domains in DNA ........................................ The systemic gene concept ................................................... 11.But… what about ‘information’? .........................................
25 27 29 30 32 35 36 49 51 52 58 58 60 61 65 67 70 73
Chapter 3. Biosemiotics and information talk in biology .... 75 Chapter 4. Information and semiosis in Peirce’s science of signs ...................................................................................... 1. Introduction .......................................................................... 2. Some basic ideas in Peirce’s semiotics ................................ 3. Kinds of Objects and Interpretants ....................................... 4. Semiosis, information, and meaning ....................................
85 87 87 91 92
Chapter 5. Some other ideas about information ................... 97 1. Information theory ................................................................ 99
10
Genes, Information, and Semiosis
2. Bateson’s understanding of information as a difference which makes a difference ..................................................... 100 3. Jablonka’s concept of semantic information ........................ 103 4. Concluding remarks ............................................................. 109 Chapter 6. A semiotic analysis of genes and genetic information: First take ............................................................ 113 1. Information talk in action: a textbook description of the ‘flow of information’ in a cell .............................................. 115 2. A general semiotic analysis of the gene as a Sign ................ 124 Chapter 7. Emergence of semiosis: A general model ........... 137 1. A multi-level approach to the emergence of semiosis in semiotic systems ................................................................... 139 2. Semiosis as an emergent process .......................................... 145 2.1. Central characteristics of emergentism: what questions about semiosis do they raise? ....................................... 148 2.2. Varieties of emergentism and questions about semiosis ......................................................................... 151 Modes of irreducibility .................................................. 151 Downward determination .............................................. 155 Unpredictability ............................................................. 162 2.3. Answering the questions about semiosis ....................... 163 What is a semiotic system? ........................................... 163 Are semiotic systems physically constituted? ............... 164 Are semiotic systems new? ........................................... 165 Is semiosis a systemic process? ..................................... 169 Is semiosis synchronically determined by the properties and arrangement of the parts in a semiotic system? ...... 170 In what sense is semiosis irreducible? ........................... 175 Is downward determination involved in semiosis? ....... 176 Is the structure of semiotic systems or processes unpredictable? ............................................................... 177 3. Concluding remarks ............................................................. 178 Chapter 8. Levels of semiosis in the genetic information system ........................................................................................ 181 1. The micro-semiotic level: strings of DNA as potential Signs .. 183
Table of contents
11
2. Transcription, RNA processing, and protein synthesis as processes of gene actualization ............................................ 186 3. Semiotic analysis of transcription ........................................ 188 4. Semiotic analysis of protein synthesis .................................. 190 5. A global picture .................................................................... 194 Chapter 9. Genes, information, and semiosis ........................ 205 1. What is genetic information? ................................................. 207 2. What is a gene? A more radical interpretation ...................... 214 3. Concluding remarks ............................................................... 221 References ................................................................................ 227 Index of names .......................................................................... 246 The authors ................................................................................ 251
Chapter 1 ‘Gene’ and ‘information’ as conceptual problems
The gene concept has certainly been one of the landmarks in the history of science in the 20th century. Gelbart (1998) and Keller (2000), for instance, call it ‘the century of the gene’. Grós (1989), by his turn, claims that we live in a ‘civilization of the gene’. Moss (2003) treats the gene as the central organizing theme of 20th century biology. Nevertheless, the definition of the term ‘gene’ remains a matter of dispute, despite the central role it plays in biological thinking. Furthermore, another concept closely related to that of a ‘gene’, namely, the concept of ‘information’, is also a matter of contention, and, in fact, continues to be an open issue in biology in general, and not only in genetics or molecular biology. The goal of this book is to contribute to the debates about the concepts of gene and information by employing theoretical and methodological tools offered by semiotics, and, in particular, by Peirce’s theory of signs. It was the Danish geneticist Wilhelm L. Johannsen who created, in 1909, the term ‘gene’, as a shortened version of de Vries’ ‘pangene’. He introduced it in an attempt to distinguish between two ideas embedded in the term ‘unit-character’, then largely used: the idea of (1) a visible character of an organism which behaves as an indivisible unit of Mendelian inheritance, and, by implication, (2) the idea of that thing in the germ-cell that produces the visible character (Falk 1986). Indeed, Johannsen was the first to be entirely successful in explicating the difference between the potential for a trait and the very trait, thanks to two other concepts he introduced, ‘genotype’ and ‘phenotype’ (Falk 1986). He made it clear that the two ideas conjoined in the term ‘unit-character’ were logically distinct, opening the door to a clear discrimination between genotype and phenotype, and, furthermore, between units in the genotype (‘genes’) and in the phenotype (‘phenes’ — a term which was largely forgotten). Genes were regarded in classical genetics as units of recombination, function, and mutation. However, it became eventually clear that genes were not units of either recombination or mutation. In the end, the prevailing meaning of the term in the 20th century was that of a gene as a ‘unit of function’. After the proposal of the double helix model and the flourishing of molecular biology, the gene was redefined as a material entity, concretely existent in DNA, and it
16
Genes, Information, and Semiosis
became usual to think of the gene also as a structural unit. Finally, the introduction of an informational vocabulary in molecular biology and genetics resulted in the so called ‘information talk’, characterized by the use of expressions such as ‘genetic information’, ‘genetic code’, ‘genetic message’, etc., and genes came to be often regarded also as informational units. Thus, a very popular notion in textbooks, the media, and the public opinion came to light, the informational conception of the gene. What is meant by ‘information’ in this case is merely sequence information in DNA or proteins (Sarkar 1998b), an idea we will challenge throughout this book. With the proposal of the double helix model of DNA by James Watson and Francis Crick, in 1953, DNA was established as the material basis of inheritance, and the road to the so called classical molecular gene concept was paved. According to this concept, a gene is a sequence of DNA that encodes a functional product, a polypeptide or an RNA. Genes seemed to be reducible, then, to concrete entities at the molecular level, namely, strings of DNA. In the molecular gene concept, the structural and functional definitions of the gene have been focused on a single entity (Stotz et al. 2004). The classical molecular gene concept is closely connected with the ‘central dogma of molecular biology’, conceived as a statement about the ‘flow’ of ‘information’ in a cell. In a manner that dramatically shows the strong reductionist tendency that marked molecular biology since its beginnings1, the very idea of the dogma was that DNA makes RNA, RNA makes proteins, and proteins make the organism (see Crick 1958). It is true that more careful renderings of the dogma could be found, as Crick himself illustrates. He also expressed the dogma as follows: “once information has passed into protein, it cannot get out again” (Crick 1958: 152–153). As expected, by ‘information’ Crick simply meant sequence information, i.e., “the specification of the amino acid sequence of the protein” (Crick 1958: 144). The ‘dogma’ became one of the elements in the hard core of molecular biology as a 1 But notice that this science seems to be moving towards the adoption of less reductionist approaches in the last years, partly due to the results of genomics and proteomics (see Keller 2000; 2005). Recently, this movement has been treated as a shift from ‘reductionism’ to ‘systems biology’ (see, e. g., Ideker et al. 2001; Chong, Ray 2002; Kitano 2002; Csete, Doyle 2002; Pennisi 2003; Stephanopoulos et al. 2004; Keller 2005; Nature 2005). Nevertheless, this is not an uncontroversial appraisal (See chapter 2).
Chapter 1. ‘Genes’ and ‘information’ as conceptual problems
17
research program. In this context, the problem that no uncontroversial conception of ‘information’ is available in biological thought becomes quite central to molecular biology. Since the beginnings of molecular biology, ‘information’ was conflated or simply identified with sequences of nucleotides in a string of DNA constituting a ‘gene’. As an example, among many others, we can quote Epp’s (1997: 537) definition of a gene as “the nucleotide sequence that stores the information which specifies the order of monomers in a final functional polypeptide or RNA molecule, or set of closely related isoforms”. When information is conceived as sequences of nucleotides in DNA, we find ourselves in a difficult position to identify other kinds of information in a cell or even in the organism as a whole. Even if we point out to other ‘informational’ molecules, such as RNAs and proteins, the ‘information’ they allegedly ‘contain’ or ‘carry’ can be directly traced down, through the central dogma, to DNA. When information is conceptualized this way, DNA becomes a sort of reservoir from where all ‘information’ in a cell flows and to which it must be ultimately reduced. Our understanding becomes, so as to say, seduced by this supposed ‘information reservoir’ and we tend, then, to overplay the role of DNA in cell systems, turning it into a complete ‘program for development’ or an all-powerful ‘controller’ of cell metabolism.2 But, as we are enchanted by this picture of the role of DNA, we simply tend to forget that DNA seems to function as a set of data rather than as a program in cell systems (Atlan, Koppel 1990). That is, we lose from sight that genes are sources of materials for cells and bookkeeping entities in evolution (Gould 2002), playing roles that are obviously important, but cannot be correctly described as master agents (or master molecules) in cell processes (Nijhout 1990). In short, we neglect the fact that it is not DNA that does things to the cell; rather, it is the cell which does things with DNA. The widespread usage of the informational conception of the gene makes the consequences of the understanding of genetic information as just sequential information go far beyond conceptual issues in 2 For critiques of this way of representing DNA and, consequently, genes (if they are regarded, as usual, as strings of DNA), see, among several other works, Oyama (2000), Nijhout (1990), Moss (1992), Smith (1994), Sarkar (1996), ElHani (1997), Griffiths, Neumann-Held (1999), Keller (2000), Guimarães, Moreira (2000).
18
Genes, Information, and Semiosis
genetics and molecular and cell biology. Oyama (2000) argues that genetic determinism is closely related to this way of interpreting what is ‘genetic information’.3 Genetic determinism, in turn, is very often an element of the ‘gene talk’ (Keller 2000), or discourse, which pervades the media and the public opinion. Through their connection with the doctrine of genetic determinism, the conceptual problems raised by the notions of ‘gene’ and ‘information’ have consequences for the public understanding of science and a whole series of social, economical, and political issues related to the knowledge and applications in the fields of genetics and molecular biology. Therefore, to clarify the concepts of gene and information in biology is an important task for both scientific and extra-scientific reasons.4 Information has been a central concept in 20th century biological thought and it is very likely that it will continue to be central in the 21st century. Ideker, Galitski and Hood (2001), in a paper about ‘systems biology’, even argue that biology is an informational science. Nevertheless, this concept is also highly problematical (see Ch. 3). Bruni (2003), for instance, successfully shows that there are several epistemological flaws in the treatment of ‘biological information’ in ‘systems biology’, most importantly, a troublesome confusion between information handled by organisms and information handled by the observer. Hoffmeyer and Emmeche (1991) argue that both the metaphors of ‘information’ and ‘program’ make the dilemma of form and substance disappear by simply treating DNA (a substance) and program (information/or potential form) as one and the same thing. Biological information, however, is not identical to genes or DNA, a view we intend to develop here. As Griffiths (2001) sums up, even a 3
In opposition to genetic determinism, Susan Oyama put forward her ‘developmental systems theory’ or ‘perspective’, in which causal parity between genes and other developmental resources is one of the most basic tenets (See Oyama 2000; Oyama et al. 2001). This perspective highlights the missing element in deterministic accounts of the genotype-phenotype relationship, namely, development, in which genes, organisms, and environments interact with each other in such a way that each is both cause and effect in a complex way (Lewontin 1983; 2000). Oyama (2000) calls this view ‘constructivist interactionism’. 4 This claim is in opposition to the idea that information-talk should be simply eliminated from biology, as advocated, for instance, by Stuart (1985) and Sarkar (1996). In this book, we argue that the concept of information and related notions play an important role in biological thinking and should not be dispensed with, but rather deepened and clarified.
Chapter 1. ‘Genes’ and ‘information’ as conceptual problems
19
widely used expression such as ‘genetic information’ is a metaphor in search of a theory (see also Oyama 2000; Stuart 1985; Sarkar 1996; Griffiths, Neumann-Held 1999). Furthermore, if Brenner (1999) is right in arguing that “the prime intellectual task of the future lies in constructing an appropriate theoretical framework for biology” — as we think he is —, then an important part of this task is to build a theory of biological information (not only genetic, but also at other levels of living systems). Nevertheless, to build such a theory and, thus, clarify the meaning of ‘information’ in biology, we should employ appropriate conceptual and methodological tools. Biosemiotics5, still a somewhat neglected perspective in current debates about the gene concept, offers a theoretical toolbox for dealing with the notion of information in biology which can help us reach a precise and coherent understanding of this central notion. We believe, in particular, that biosemiotics makes it possible to formulate the notion of genetic information in a manner which does not lend support to genetic determinism. As regards the gene concept, several discoveries in molecular biology, including transposons, split genes, alternative splicing, consensus sequences, overlapping and nested genes, mRNA editing, transplicing, etc., posed very difficult problems to the generic or consensus view of genes6, much in line with the classical molecular gene concept. These discoveries led, in Falk’s (1986: 164) words, to “... an age of anarchy in the instrumental formulation of genetic entities”, in which a great number of heterodox entities was admitted into the “expanding zoo of genetic units”. In this scenario, the question ‘What is a gene?’ became a topic of strong debate in the philosophy of biology. Recent advances in molecular biology and its daughters, genomics and bioinformatics, brought to the surface a swarm of new problems. Symptomatically, we find doubts about the status of the gene concept not only in papers by philosophers of biology, but also in empirical 5
For a historical introduction to biosemiotics, see Kull (1999) and Favareau (2007). For an introduction to basic ideas in biosemiotics, see an introduction in Hoffmeyer (1996), and a more comprehensive treatment in Hoffmeyer 2008a. Another introduction is Emmeche, Kull, Stjernfelt (2002). See also non-Peircean accounts of biosemiotics in Barbieri (2003) and Markoš (2002). 6 See Falk (1986), Portin (1993), Fogle (1990, 2000), Pardini, Guimarães (1992), Sarkar (1998a), Griffiths, Neumann-Held (1999), Keller (2000), Pearson (2006). See also Chapter 2 in this book.
20
Genes, Information, and Semiosis
papers inside molecular biology (possibly indicating a crisis in molecular biology as a ‘normal science’, in a Kuhnian sense). In Ch. 2, we will discuss some examples of empirical papers in which the troublesome status of the gene concept is explicitly recognized. As 20th century came to a close and we entered in what seems to be a whole new era in biological research, the future of the gene didn’t look bright for some thinkers. Keller (2000), for instance, considered the gene a concept ‘in trouble’ and suggested that maybe the time was ripe to forge new words and leave that concept aside. She was inspired by similar ideas put forward by Portin (1993) and Gelbart (1998). Portin (1993) claimed that our knowledge about the structure and function of the genetic material has surpassed the terminology typically used to describe it, and it may even be the case that the term ‘gene’ is not useful anymore. In turn, Gelbart (1998) argued — in a rather naïve realist tone — that genes are not ‘physical objects’ but merely ‘concepts’, which played an important role in previous times, but had turned out to be of limited value and might in fact be a hindrance to our understanding of the genome, since genome organization is much more complex than can be accommodated in the classical gene concept. Although some authors agreed with Keller (e. g., Rios 2004), her proposal has not found wide acclaim; rather, it was rejected by many reviewers of her book, such as Coyne (2000), Magurran (2000), Maynard Smith (2000b), Hall (2001), Wilkins (2002), and Moyle (2002).7 Recently, she published a paper reexamining her previous ideas under the light of developments in molecular biology, genomics, and related areas since The Century of the Gene appeared (Keller 2005). In this paper, she takes a more optimistic view about the future of the gene, considering her previous arguments for the need to move on to a “century beyond the gene” (Keller 2005: 3) an “impassioned” one, and arguing that the 21st century will be the century of genetic systems, rather than of the gene.8 She does not claim anymore that the 7
In this connection, we can also quote Judson’s (2001: 769) claim that “… we cannot abandon the term gene and its allies”, even though he doesn’t mention Keller’s book. 8 She stresses that when one tries to be “a historian of the present” one faces many pitfalls, and the most serious is “that history can happen a lot faster than a scholar […] can write” (Keller 2005: 3). In fact, as Moyle (2002) reminds us, we can say that Keller walked her way through slippery grounds when claiming
Chapter 1. ‘Genes’ and ‘information’ as conceptual problems
21
gene concept should be abandoned, but rather considers that the challenges currently posed by biological complexity demands “new ways of talking” (Keller 2005: 8–9). In order to address the interactions between the parts of living systems and the dynamics of these interactions, biologists should overcome, Keller argues, “ingrained habits of thought and speech that give ontological priority to those parts”. These habits are particularly problematic in genetics, “where the parts are taken to be genes”, while “genes, by definition, do not have meaning in isolation”. Keller goes on to treat the cell as “a meaning making system that turns nucleotide sequences into genes”. In this picture, the gene concept can survive in the 21st century, she argues, but “by reconceptualizing them as verbs” (Keller 2005: 9). It is interesting to see Keller’s comments in this recent paper under the light of the results of the semiotic analysis developed in this book, since, in one interpretation of them, genes are treated as processes, rather than entities or even sequences of entities. Thus, they are reconceptualized as verbs, rather than nouns, or to be more precise, as a special kind of complex signs, similar to linguistic signs but not to be confused with the particularities of language9 and rather analyzed by means of general semiotic concepts (see Ch. 9). As even a brief inspection of textbooks used in molecular biology or genetics courses will show, with their ever-growing chapters on cell signaling, “communication has become the new buzz word in biology”. After all, it is now widely recognized that biological function is not found in genes, DNA or protein products considered in isolation, but rather “in the communication networks of which the DNA and the proteins are part” (Keller 2005: 9). It is remarkable how Keller’s eloquent statements clearly show the need of biosemiotic conceptual and methodological tools for understanding how cells turn DNA sequences into genes! And, as she writes, communication is just one term. What we need is more than this, we need a conceptual framework to deal with communication and information in living systems in a more adequate way, and we truly believe that biosemithat the ‘gene’ was (almost) gone. History was indeed faster than her. But this is not surprising, and it is nice to see her willingness to reappraise her ideas in view of changes in research strategies and knowledge in molecular biology and related areas. 9 See Emmeche and Hoffmeyer (1991) for an early argument that a strict analogy between genetic and linguistic signs cannot be maintained.
22
Genes, Information, and Semiosis
otics has something important to offer to the efforts to build such a framework. Other philosophers of biology and also practicing scientists take a more optimistic view about the future of the gene concept. Falk, for instance, while admitting that the gene is a concept ‘in tension’ (Falk 2000), seeks ways to ‘save’ it (Falk 2001). Waters is even more optimistic, considering that the problem of conceptual variation and ambiguities concerning the term ‘gene’ can be overcome by unifying different definitions of the gene by building a concept with a number of ‘open’ clauses, such as that of “a gene for a linear sequence in a product at some stage of genetic expression” (Waters 1994: 178). Hall (2001) is also optimistic, arguing that, despite published obituaries10, the gene is not dead, but alive and well, even though ‘orphaned’, ‘homeless’, and seeking a haven from which to steer a course to its ‘natural’ home, the cell as a fundamental morphogenetic unit of morphological change in development and of evo-devo (the interface between evolution and development). Kitcher (1982: 355–356) does not think of the conceptual variation regarding what is a gene as necessarily a problem. For him, we can allow the term ‘gene’ to have such a heterogeneous reference potential, since different gene concepts may be useful for different areas of research, satisfying different purposes (see also Burian 1985). The attempts to save the gene indeed led to distinctions between different concepts, as, for instance, Moss’ (2001, 2003) distinction between gene-P (the gene as a determinant of phenotypes or phenotypic differences) and gene-D (the gene as a developmental resource).11 In this book, we will be specifically interested in genes as conceived from the perspective of molecular biology (roughly Moss’ gene-D). Our task here is to begin the construction of a theoretical framework for a semiotic analysis of the concepts of ‘gene’ and ‘information’ on the grounds of a case study about protein-coding genes.12 We should emphasize the originality of this approach, not 10
Hall (2001) mentions Gray (1992), Neumann-Held (1999), and Keller (2000). See Ch. 2 for more details and other distinctions between different gene concepts. 12 This is only a methodological decision, made for the sake of simplicity. We intend to build a theoretical framework for future analyses on the grounds of the more well-established case, subsequently applying it to more complex and difficult cases. 11
Chapter 1. ‘Genes’ and ‘information’ as conceptual problems
23
only in the specific context of molecular biology, but also in the general context of biosemiotics. We think it is important to develop biosemiotics by providing new sets of ‘modeling tools’ and some exemplars or case studies to understand the precise sense in which specific life processes can be conceived as involving the action of signs, as generally claimed by biosemioticians. Furthermore, by applying the formal notion of semiosis to model some aspects of the genetic system, we intend to produce a radically new explanation of ‘genetic information’ as a semiotic dynamical process. Thus, we move towards a reinterpretation of what is information in a cell that hopefully avoids a number of problems detected in information talk not only in biology but also in science as a whole.
Chapter 2 The problem of the gene
1. The birth of the ‘gene’ as an instrumental concept The ideas underlying the concept of a ‘gene’ can be traced back to Gregor Mendel’s use of the German words ‘charakter’, ‘element’, ‘faktor’, and ‘merkmale’ as means of describing the determinants of particulate inheritance. Nevertheless, the term itself was created in the beginning of the 20th century. After 1900, as research on genetics increased rapidly, a consensual technical term designating the potential for a trait became more and more necessary. This term was provided by the Danish geneticist W. L. Johannsen who proposed, in 1909, the term ‘gene’, as a way of clarifying the ideas involved in the term ‘unit-character’, something Johannsen was in a position to do due to his successful explanation of the difference between ‘genotype’ and ‘phenotype’ (see Ch. 1). In the beginnings of genetics, the ‘gene’ was often regarded as a useful abstract concept to express regularities in the transmission of phenotypic traits which had, however, a highly uncertain material counterpart. To put it differently, an instrumentalist view about the status of ‘gene’ as a theoretical concept prevailed (Falk 1986). Mendel treated his ‘factors’ in an instrumentalist way, as just a useful accounting or calculating unit in his experiments. Johannsen himself had similar ideas about the term ‘gene’, seeing it as a very handy term but with no material counterpart that could be related to it with any degree of confidence (Johannsen 1909, cited by Falk 1986). Although he accepted that heredity was based on physicochemical processes, he warned against the conception of the gene as a material, morphologically characterized structure, claiming that “the gene is [...] to be used as a kind of accounting or calculating unit [Rechnungseinheit]’ (Johannsen 1909, cited by Falk 1986). In 1909, Johannsen, “was quite uncertain as to the nature of the genes, having probably even no idea that genes could be described as real chemical substances; they seem to have been mere ‘calculation tablets’ to him” (Wanscher 1975: 135). Falk (1986: 140–141) warns, however, that even though Johannsen’s attitude towards the gene was instrumentalist, he also played with the idea of ‘something’ (Etwas) that was present in the gametes and the zygote, he talked about ‘foundations’ and ‘determiners’ of
28
Genes, Information, and Semiosis
inheritance. Anyway, he insisted that his terminology was not committed to any specific conception about the nature of the gene. For him, no hypothesis about the nature of that ‘something’ could be supported. The term ‘gene’ was meant to express merely the fact that many characteristics of an organism are specified in the gametes by ‘special conditions’, ‘foundations’, and ‘determiners’ present in ‘unique’, ‘separate’ and ‘independent’ ways (Johannsen 1909, cited by Falk 1986), i.e., as units of particulate inheritance. In short, for Johannsen ‘gene’ was a theoretical concept, reached by inference, to which no hypothesis about the existence of the entities thus named could be added. The reason why Johannsen adopted this mostly instrumentalist attitude is clearly an outcome of the state of knowledge in the beginnings of the 20th century. The ‘gene’, named by Johannsen less than ten years after the rediscovery of Mendel’s laws, was then defined with reference to a specific constant relationship between observable characters at the level of the phenotype. That is, a gene (that ‘something’ which was the potential for a trait) could only be recognized by its ‘representative’, the trait, or, to be more precise, by the alternative appearances of the trait. Observed characters or traits were only the ‘markers’ for ‘genes’, which had, in fact, to be inferred (Falk 1986: 144). This situation made any ascription of a clear and definite meaning to the material counterparts of genes very difficult, maybe even impossible. It is easy to understand, then, why Johannsen was so cautious as regards any hypothesis about the material reality of genes, but, at the same time, made allusions to ‘something’ that might be present in the gametes and the zygote as a putative ‘determiner’ of traits. Many early Mendelians stated that genes were molecules, but this was nothing but a bold speculative hypothesis before the 1940s. The gene concept appeared in the context of a position which Falk (1986: 141) calls ‘instrumental reductionism’, an approach which was very fruitful for the development of genetics, as the success of T. H. Morgan’s school shows. Morgan instrumentally manipulated genes ‘as if’ they were material units, even though he was not deeply concerned with the nature of these units, since he didn’t think it would make the slightest difference to genetic experiments whether the gene was a hypothetical unit or a material particle (Falk 1986).
Chapter 2. The problem of the gene
29
With the growth of knowledge in Mendelian genetics, and through a series of developments beyond the scope of this chapter (such as the building of Morgan’s chromosome theory of heredity and advancements in the understanding of the physicochemical basis of the genetic material, as well as of the relationship between genes and proteins. See, e. g., Carlson 1966; Kitcher 1982; Mayr 1982; Falk 1986; Fogle 1990; Portin 1993; Keller 2000), this instrumentalist attitude was superseded by a material understanding of the gene. A notorious member of Morgan’s group, Herman J. Muller, was one of the first advocates of the latter conception, which ultimately gained wide acceptance throughout the 20th century, mainly after the proposal of the double helix model of DNA by Watson and Crick.
2. The fruitful tension between instrumentalist and realist views of the gene Muller was one of the main supporters of the idea that genes were material units in their own rights, even though they could only be then recognized through their effects. Even though he was in Morgan’s group, Muller’s attitude towards the gene was considerably different from Morgan’s. As Falk (1986) comments, while Morgan stressed that the gene was determined through its phenotype (clearly showing a tendency to treat genes as instrumental constructs), Muller insisted that the gene determined the phenotype, visibly treating it as a material entity which acted as a causal factor. Muller (1922, quoted by Falk 1986) resolutely stated that genes existed as “ultra-microscopic particles” in the chromosomes, arguing against their description as related to “a purely idealistic concept, divorced from real things” (Muller 1947, quoted by Falk 1986). As Falk (1986: 150) stresses, the gene has now been reified. In such a realist view, it was not seen as an instrumental fiction, just a useful accounting or calculating unit. Muller’s material conception of the gene contributed to the establishment of a biological setting for the coming chemical attack on the nature of the gene as well as on the mechanism of its duplication, mutation, and functioning. In the first decades of genetics, the gene concept evolved mainly through the competition of the instrumentalist and realist views. Falk (1986) argues that this evolution reflects a dialectical confrontation, a
30
Genes, Information, and Semiosis
constant mutual interchange between the instrumental and the material approaches to the nature of the gene. This tension resulted in a development in a ‘Russian dolls’ pattern: the elaboration of a functional meaning leading to the investigation of a deeper structural meaning, and this leading, in turn, to a still deeper level of functional meaning, and so on. But there were important dissenters. Goldschmidt, for instance, proposed a ‘holistic’ view about the basis of inheritance which was incommensurable with both the instrumental and the material views (Falk 1986: 154–157). For Goldschmidt, genes were abstractions that had been helpful in organizing observations at the early days of genetics, but they were not helpful more. He argued against Muller’s reductionist philosophy, denouncing it as ‘hyperatomistic’, and claiming that it introduced unproved systems of units for more and more genetic permutations, while losing from sight that the whole chromosome was the ‘real’ unit of structure. Even though experimental data pointed to ‘loci’ along chromosomes, adding to the observational data any assumptions about the reality of these loci was, for Goldschmidt, an instance of hyperatomism.
3. Genes as units in Mendelian genetics Since its beginnings, Mendelian genetics was committed to the postulation of a one-to-one correspondence between a gene and some developmental unit (Griffiths, Neumann-Held 1999). Accordingly, the gene was conceived as a unit of (1) function, (2) mutation, and (3) recombination (Mayr 1982: 795–796). In Mayr’s words, this entailed a ‘bean-bag’ view of the genotype, according to which each gene is independent in its actions and in the effects of selection on it. The treatment of genes as ‘units’ faced increasing difficulties as genetic research advanced in the 20th Century. Many of these problems came to light in the last three decades, but already in 1925 the first counter-evidence was provided by Sturtevant’s discovery of position effect (Mayr 1982: 797–798). Studying the ‘bar’ gene in Drosophila melanogaster, Sturtevant showed that the position of this gene in the chromosomes could change its effects on the phenotype. Thus, the function of a gene and its phenotypic effect could be modified merely by altering the arrangement of genes in chromosomes, in
Chapter 2. The problem of the gene
31
the absence of mutation or any change in the quantity of genetic material. Position effect showed, thus, that there were at least difficulties in treating the gene as a unit of function. Sturtevant’s studies of the ‘bar’ gene in D. melanogaster also provided evidence for unequal crossing over, which was subsequently shown to be much more common than initially expected. Therefore, the unit of recombination was not necessarily the gene. Mutational analysis, in turn, showed that several different mutational sites could be found within a single gene. Consequently, the gene was not the unit of mutation. Thus, the idea that the gene could be simultaneously a unit of recombination, mutation, and function did not hold, and, in the end, the idea that prevailed was that of a gene as a ‘unit of function’, despite position effect. Benzer’s (1957) work on the fine structure of genes played an important role in the establishment of the gene as a unit of function. Benzer showed that units of function (in his words, ‘cistrons’, a term often used in the primary literature in the place of ‘gene’) are typically much larger than units of recombination (which he called ‘recons’) and units of mutation (which he called ‘mutons’).13 Kitcher (1982) shows how Benzer, although proposing, quite radically, that the term ‘gene’ could be abandoned altogether, being replaced by the terms ‘cistron’, ‘muton’ and ‘recon’, in fact refined the very concept of the gene, by means of his cis-trans test (for details, see Benzer [1957]). In a scenario where the development of classical genetics had resulted in a heterogeneity in the reference potential of ‘gene’, casting doubt on some of the assumptions with which the term was laden, Benzer accommodated, with his concept of cistron, many of the references of his predecessors, while modifying the very meaning of the term ‘gene’. Although the conception of genes as units of function prevailed in classical genetics, with the advent of molecular biology two other ideas were added to our understanding, namely, those of genes as units of structure and of information (see below). 13
The terms ‘muton’ and ‘recon’ were deleted from the vocabulary of genetics in virtue of the belief, after the advent of molecular biology, that Benzer’s units of mutation and recombination were single nucleotides along the DNA molecule. But this was premature, since neither the unit of recombination nor that of mutation is a single nucleotide (with the exception of point mutations) (See Falk 1986: 158).
32
Genes, Information, and Semiosis
4. The classical molecular gene concept In the first half-century of genetics, the understanding of the material structure of genes showed remarkably little advance. But the historical development of Mendelian genetics helped prepare the way for what is often regarded as a scientific revolution brought about by the advent of molecular genetics. The concept of the material gene, advocated by Muller, among others, resulted in a deeper insight into gene function and structure, and played a role in the process which led to the proposal of the double helix model of DNA. Watson and Crick’s (1953) model of the DNA structure was undoubtedly an astounding achievement of the 20th century biology. Understanding the structure of DNA was, in fact, a requisite for grasping its role in inheritance, development, and cell metabolism. The model for the structure of DNA proposed by Watson and Crick provided at one stroke a physicochemical understanding of a whole set of requirements for the material gene, as established in the conceptual framework of classical genetic research: It explained the nature of the linear sequence of genes; suggested a mechanism for the exact replication of genes (as well as for the synthesis of RNA from DNA sequences); chemically explained the nature of mutations; and showed how mutation, recombination, and function are separable phenomena at the molecular level.14 It was mainly the proposal of an acceptable model for the structure of DNA that made the gene truly emerge as a concrete spatial elementary entity of heredity (Keller 2000). It amounted, in short, to a triumph of the realist over the instrumental 14
The proposal of the double helix model gave birth to the research program of molecular biology. The outstanding development of this branch of biology, in turn, brought about the difficult question of what is the relationship between Mendelian and molecular genetics. Since Schaffner (1967), philosophers of science have been discussing the possibility that the relationship between Mendelian and molecular genetics, as successive knowledge structures in the same field, is such that the former has been absorbed by, or theoretically reduced to, the latter. Several authors played an important role in this debate, such as Hull (1974), Maull (1977), Wimsatt (1976), Kitcher (1982), among others. With regard to the outcome of these debates, what we can say is that it is at least unclear that the relationship between these two sciences can be explained as one of reduction (for a thorough discussion, see, e. g., Sarkar 1998). Anyway, we will not address here the debate about a putative epistemological reduction of Mendelian to molecular genetics, as this issue is outside the scope of this book.
Chapter 2. The problem of the gene
33
view of the gene, establishing DNA as the material basis of inheritance. This model was the basis for the so called classical molecular gene concept, according to which a gene is a stretch of DNA that encodes a functional product, a single polypeptide chain or RNA molecule. In this concept, a gene is treated as an uninterrupted unit in the genome, with a clear beginning and a clear ending, which performs one single function. It is therefore a concept of both a structural and a functional unit in the genome. In these terms, the material understanding of the gene structure provided by Watson and Crick’s model and the wave of molecular genetics research that followed allowed genetics to overcome for some time, at least, a dialectical tension that has pervaded its history, namely, between the gene as operationally identified by its product and the function it plays, and the gene as a segment in its own right, independently of whether it is transcribed, and whether its transcript is translated (Falk 1986: 159). Genetics was committed, from the very beginning, to the idea that a gene is some kind of chromosomal segment. Nonetheless, this is not so simple an idea as it may seem at first, since it entails the segmentation problem, i.e., the problem of how to divide chromosomes into genes in an appropriate way. In the evolution of the gene concept, two basic approaches to this problem played an important role (Kitcher 1982). In one approach, genes were to be found in chromosomes through their functions in the production of phenotypic effects. In another, genes should be identified by means of their immediate action. The first approach was characteristic of classical genetics, the second, of molecular genetics. Therefore, in molecular genetics the tension between a functional and a structural view of the gene remained, but the classical molecular gene concept contribute to dissolve this tension by bringing a structural dimension to the, until then, predominantly functional view of the gene as a unit. As Kitcher (1982: 356) sums up, “molecular biologists, when pressed to identify the referent of ‘gene’, have adopted, almost universally, the view that genes (more exactly structural genes) are chromosomal segments which code for particular polypeptides” (see also Burian 1985). By bringing together the structural and functional definitions of a gene, the classical molecular gene concept showed substantial
34
Genes, Information, and Semiosis
explanatory, predictive, and heuristic powers: the molecular gene initially had a well-defined structure, with easily determinable beginning and end, a singular function, and an easily understandable mechanics. The classical molecular concept updated the Mendelian particulate model and the related interpretation of genes as units, by treating the gene as a discrete region of DNA with simple organization and unitary function. In these terms, the molecular understanding of what a gene is was superimposed onto the Mendelian idea of ‘unit’ (Fogle 1990). As Fogle (1990: 352) comments, “the physical gene elaborated in the molecular era was something with shape, defined structure, and comprehensible mechanics. Following at the heels of the bead-on-a-string model, it was an extension of the particulate entity arrayed onto chromosomes. The physical gene was therefore a linearly organized set of instructions that gave rise to phenotypic expression”. In the classical molecular gene concept, genes could be convincingly treated, therefore, as both functional and structural units. And, with the introduction of an informational vocabulary in molecular biology and genetics (see Kay 1995, 2000), genes came to be often regarded also as informational units, leading to what has been called the informational conception of the gene, a very popular notion in textbooks, media, and public opinion. The classical molecular gene concept is also closely connected with the ‘central dogma of molecular biology’, as we mentioned in Ch. 1. This dogma was framed in terms of the flow of ‘information’ in the cell. But what is meant by ‘information’ in this case is merely sequence information in DNA or proteins (Sarkar 1998b). As we will argue in the next chapter, it is not clear at all that this is an adequate definition for ‘genetic information’. Genetic determinism is also closely related to this interpretation of ‘genetic information’; the univocal correspondence between gene/protein/phenotypic trait postulated in Mendelian genetics and updated by molecular biology can be seen as one of the fundamental grounds of genetic determinism (Leite 2005). In Fogle’s view, it was the attempt to keep and even save the idea of genes as units of structure and/or function (or, for that matter, and/or information) that led to two aspects of the now largely recognized crisis of the gene concept: the proliferation of meanings ascribed to the term ‘gene’ and the failure in acknowledging the diversity of gene architecture, particularly in eukaryotes. It is true that this
Chapter 2. The problem of the gene
35
updating of the Mendelian view of genes as units began before the double helix. In the 1940s, Beadle and Tatum (1941) concluded, from their investigations about the nature and physiological function of the gene, that one gene would correspond to one ‘primary’ character and one enzyme. But this was only the beginning. The classical molecular gene concept involved, in fact, a reinterpretation of the ‘one gene-one enzyme’ hypothesis, which turned it into the ‘one gene-one polypeptide’ and then ‘one gene-one polypeptide or RNA’ hypotheses. It is clear, however, that this reinterpretation had no deep consequences for the unit concept itself. Nevertheless, the coherent relationship between genes at the molecular level and Mendelian entities, at first successful, would not survive the increasing understanding of the architectural diversity of the molecular gene, as we will see below.
5. The golden years of molecular genetics Crick’s (1958) suggestion that amino acid sequences of proteins were the most delicate expression of an organism’s phenotype paved the way to a deeper understanding of the nature of the ‘gene as a unit’. In his 1958 landmark paper, ‘On protein synthesis’, Crick suggested the sequence hypothesis, namely, the hypothesis that the relation between genes and proteins could be explained by means of the correspondence between sequences of nucleotides in DNA and sequences of amino acids in proteins. Evidently, this hypothesis is strictly connected with the central dogma, presented in that very same paper. As any biology student knows, the sequence hypothesis became part of common knowledge in molecular biology. It was a simple but ingenious idea: the specificity of a piece of nucleic acid would reside only in the sequence of bases, and that sequence would be a simple code for the sequence of amino acids in a given protein. The ‘golden rush’ to decipher the genetic code had begun: inspired by the sequence hypothesis, molecular biologists ended up deciphering the genetic code in 1966, less than ten years after Crick suggested his hypothesis. Those were times of great confidence for geneticists and molecular biologists. They had elucidated the nature and dynamics of the genetic material, established the rules of the genetic code, solved the riddle of protein synthesis, showing how sequence ‘information’ in the genetic
36
Genes, Information, and Semiosis
material could serve as a basis for that process, and so on. Stent (1968), for instance, stated that the major problems of molecular genetics were all solved and geneticists would only have to iron out the details. It is easy to see the irony in this story: Stent wrote that just some years before the development of genetic engineering techniques, which brought about a wealth of new data on genetic systems, with deep consequences to our understanding of those systems, and not only at the level of the details. Despite all the tensions embodied in the classical molecular gene concept, the first difficulties faced by it, in those times of great confidence, were easily accommodated. For instance, the discovery that some segments in DNA are transcribed but not translated (rRNA, tRNA, and other RNA genes) and others are not even transcribed (‘pseudogenes’) didn’t seem to affect, at first, the concept of the material gene. It was equally easy to adjust the understanding of gene function so as to do justice to the discovery that proteins could be composed of several polypeptides of genetically independent origin: the slogan of the gene simply became ‘one cistron-one polypeptide’. And, again, there was no problem in extending the concept of the gene to encompass the difference between structural and regulatory genes (for details, see Keller 2000).
6. How the gene concept became a problem: why is the gene not a structural unit? As our knowledge about the genetic material increased, particularly of eukaryotes, the structure and boundaries of molecular genes became less and less clear. Many of these problems seriously affected the idea that genes could be treated as structural units in DNA. The problems with the gene concept can be explained as a consequence of three features, established by molecular biology/genetics: (i) one-to-many correspondences between DNA segments and RNAs/polypeptides (e. g., alternative splicing); (ii) many-to-one correspondences between DNA segments and RNAs/polypeptides (e. g., genomic rearrangements); and (iii) lack of correspondence between DNA segments and RNAs/polypeptides (e. g., mRNA editing). The boundaries and classifications of genetic units began to dissolve by the time Jacob and Monod (1961) proposed the operon model for
Chapter 2. The problem of the gene
37
gene regulation (Falk 1986; Fogle 1990; Keller 2000). In this model, a new feature was the distinction between structural and regulatory genes. But, as gene regulation involves the synthesis of proteins, and, consequently, there were at first no difficulty in treating regulatory genes as structural and functional units, this feature didn’t really challenged the classical molecular gene concept. A harder problem concerns, however, the fact that in bacterial operons one single polycistronic mRNA is produced and its sequential translation results in different polypeptides (Jacob, Monod 1961). In the lac operon, for instance, three structural genes and a regulator gene can be seen as a message unit which results in a product. Or, to put it differently, it was not clear in the lac operon how to identify the units involved, i.e., whether the unit comprised the set of four genes functioning together, or each gene in itself could be treated as a separate unit.15 There were, however, even harder problems in the operon model, as that of the status of the operator and promoter regions (Falk 1986; Fogle 1990; Keller 2000). They have functions, are heritable, undergo mutations, and influence the phenotype, but despite sharing all these properties with genetic units, they are not regarded as genes. The reason for this is simply that they are not transcribed. Consequently, they turned out to be called the operator ‘region’ and the promoter ‘site’. The expansion of the ‘zoo’ of genetic entities (Falk 1986), which would in time become a severe problem for our understanding of genomes, had begun. As Fogle (1990: 352) summarizes, this case shows in a clear way that the material gene concept has a restricted meaning in the jargon of genetics and doesn’t apply equally to all genomic regions which affect the phenotype. But how can we justify this dyadic system of ‘units’ (genes) and ‘non-units’ (non-transcribed functional sites, repetitive DNA etc.), if they share so many properties? It was the juxtaposition of the classical and molecular perspectives in the molecular gene concept that led to this hardly justifiable and rather arbitrary dyadic system. As Fogle (1990: 368) writes, “layering molecular genetics 15 This problem is not at all restricted to prokaryotes. In eukaryotes, we find translation products which amount to tandemly joined polypeptides and are split into several products through proteolytic cleavage. An example is offered by proopiomelanocortin, in which various tissue-specific products are encoded in a single exon (Pardini, Guimarães 1992).
38
Genes, Information, and Semiosis
over classical genetics ignores much of the molecular complexity because only two conventional categories of phenomena are permitted: either something is a gene or it is something else (repetitive DNA, regulator sequences, pseudogene, etc.)”. The operon model also entailed a transition from the idea of gene action to that of gene activation, according to which it is not the case that all genes are constitutively expressed, but rather many genes can be turned on or off depending on the context in which a cell is embedded (Keller 2000). The idea of gene activation also stressed how the activity of a gene depends on elements outside the coding sequence itself (and, later, it became clear that this dependence had far-reaching consequences for evolutionary processes, since in many cases the coding sequence of genes is conserved, while changes in the regulatory elements make the same gene products be present at different times and places in an organism). Another problem resulted from Britten and Kohne’s discovery (1968) that eukaryotic chromosomes show regions of highly redundant DNA, but it was also easily accommodated by the molecular gene concept. In fact, much of the eukaryotic DNA is highly repetitive, appearing in up to a million repeats of the same sequence, often with very low informational content. These sequences posed the problem of whether we can say that there are genes there (Falk 1986). Highly repetitive sequences such as rRNA genes, which are simply conventional genes present in many copies, can be easily interpreted in accordance with the classical molecular gene concept. But when we consider sequences with low informational content repeated a thousand times in the genotype, it is quite clear that we would be stretching the gene concept too much or even misusing it if we intended to call these sequences ‘genes’. They may have been preserved by natural selection because they fulfill ‘household functions’, playing a role in the maintenance of chromosome structure, but, even so, would it be appropriate to call every sequence of DNA that fulfills some function a ‘gene’? If we did so, couldn’t we be running the risk of turning this term into an empty sign with an excessive capacity for signification? But, if repetitive sequences in DNA are not genes, what are they? Things got even more bewildering with the discovery that insertion segments with neither fixed sites nor constant numbers are ubiquitous in the genome. Subsequently, transposons were discovered, i.e.,
Chapter 2. The problem of the gene
39
insertion segments that carry one or more genes, defying the established notion of genes located at specific sites along the chromosomes. Even these anomalies, however, were accommodated by the classical molecular gene concept. Anyway, even though these first incongruities were accommodated by this concept, anomalies gradually accumulated to such an extent that it was not possible anymore to avoid pondering about the validity of that standard interpretation of genes. To understand how the gene concept became a problem, it is worth considering, first, difficulties faced by the idea that a gene is a structural unit in the genome, and, subsequently, move on to the interpretation of the gene as a functional unit. Fogle (1990) examined four possible structural models for a protein-coding gene (see Figure 1). Model A includes the transcribed region and all neighboring sequences with detectable influences on gene expression. Model B is limited to the transcribed region. Model C includes only the set of exons derived from a pre-mRNA. Finally, model D is limited to the coding exons of a primary transcript, excluding non-coding leader and trailer sequences. Model A is the most inclusive, incorporating all cis-acting sequences which influence transcription, such as promoters, enhancers, terminators, regulators, etc. This model has the advantage of breaking with the arbitrary distinction between units and non-units in the genome. Nevertheless, it faces a host of problems which arguably make its costs overcome its benefits. Most of these problems are related to the fact that there are many different types of regulatory elements, generally operating in complex and varied combinations. There are cis-acting factors which influence transcription independently of their distance from the coding sequences, such as enhancers and silencers, making it difficult to empirically assign the boundaries of a gene. There are cis-acting factors which simultaneously affect the expression of different genes. There are even cis-acting factors which are nonspecific, influencing any compatible promoter within their range (Fogle 1990). Therefore, model A will lead to substantial overlapping of many genes which depend on the same regulating sequences, raising difficulties to the idea that a gene is a structural unit.
40
Genes, Information, and Semiosis
Figure 1. Four structural models for the protein-coding gene (Fogle 1990). The solid lines represent areas included in each model.
Consider, also, that, in order to justify the inclusion of a cis-acting sequence in a gene, it is only necessary to show that it modulates transcription. This leads to important problems as regards the decision about including or excluding such sequences from a gene. Let us consider, for instance, position effects: if one shows that the fact that a given rearrangement of genetic material ends up placing a gene near heterochromatin (an event which is not uncommon), and the expression of the gene is significantly affected by this position, model A will demand that a huge part of a chromosome be included in the gene. After all, its transcription will be modulated by a whole region of heterochromatin.16 Another problem concerns the fact that cooperation 16 Interestingly enough, position effects were one of the main grounds for Goldschmidt’s critique of Muller’s ‘hyperatomistic’ view, in which genes were taken as material units in chromosomes, and for his argument for a ‘holistic view’, according to which the whole chromosome was the unit of structure in the genetic material (see above).
Chapter 2. The problem of the gene
41
between cis-acting factors make it difficult to decide about including or excluding these sequences from a gene, even when we are not dealing with such extreme cases as position effects. Moreover, the decrease in detectable influence of cis-acting sequences as they are located farther and farther upstream of the gene puts into question where the border of such an all-inclusive gene would be. Finally, there is already a term largely accepted for a unit of gene expression, namely, ‘operon’, and, as Epp (1997) argues, to include regulatory sequences in genes could lead to a conflation with operons. All these problems, among several others, suggest that we have to abandon a completely inclusive model for the structural gene (Fogle 1990). The next model to consider, then, is model B, in which the structural boundaries of the gene are defined by the process of transcription. This model is appealing, since it relies on the clear borders that transcription seems at first to establish, and, moreover, supports an interesting relationship between a transcription unit and the sequences necessary to make a polypeptide. Nevertheless, it is challenged by two particularly troublesome phenomena, split genes and alternative splicing. Discovered in the end of the 1970s, split genes contain coding (exons) and non-coding (introns) regions, posing a first problem to model B, since it relies on the transcription unit to demarcate what is a gene. As introns are excised during RNA splicing and the coding exons are combined to form a mature, functional mRNA, the sequences transcribed into RNA are not the same as those later translated into proteins. Therefore, a protein coded by a spliced mRNA molecule exists as a chromosomal entity only in potential (Keller 2005). The situation becomes more perplexing, and less promising with regard to the prospect of delimiting genes as entities to which we can ascribe a single, well-defined transcript, when we consider the diversity of splicing patterns of the same primary transcript, a phenomenon named ‘alternative RNA splicing’. The vast majority of genes in multicellular eukaryotes contain multiple introns and the presence of such introns allows the expression of multiple related proteins (isoforms) from a single stretch of DNA by means of alternative splicing (e. g., Black 2000). The fibronectin (FN) gene offers a typical example. Fibronectin is a mammalian extracellular adhesive protein which plays several important functions (Kornblihtt
42
Genes, Information, and Semiosis
et al. 1996; Miyamoto et al. 1998; Danen, Yamada 2001). The FN gene is an interesting example because it has become paradigmatic to illustrate not only the generation of protein diversity by alternative RNA splicing, but also several other biological phenomena, such as genome evolution and topological coordination between transcription and splicing (Kornblihtt et al. 1996). This gene contains numerous exons, and, after the FN pre-mRNA is transcribed from DNA, it undergoes cell type-, development- and age-specific splicing. More than 20 different FN isoforms have been identified. Each of these isoforms is encoded by a differently, alternatively spliced mRNA, and, therefore, each isoform results from a unique combination of exons found in the FN gene. Alternative splicing makes model B, and, generally speaking, the whole idea of genes as units (no matter if structural or functional) very clumsy: alternatively spliced pre-mRNAs give rise to several different gene products from the same stretch of DNA.17 Alternative RNA splicing requires that the conceptualizations of genes move far beyond the simple scheme captured in formulas such as ‘one gene-one protein or polypeptide’. The challenge of alternative splicing might be accommodated, however, by simply replacing this formula by a new one, for instance, ‘one gene-many proteins or polypeptides’. But the situation doesn’t seem so simple. First, because genes do not autonomously ‘choose’ splicing patterns, they would lose a substantial aspect of their specificity, with respect to the polypeptides which will be synthesized. Splicing patterns are subject to a complex regulatory dynamics which, after all, involves the cell as a whole (Keller 2000). Second, given that the segment of DNA which is transcribed as one unit into RNA is translated into several distinct polypeptides, the question ‘Which segment is, at last, the gene?’ cannot be avoided. On the one hand, the unit transcribed into a single RNA could count as a gene, i.e., we might call a ‘gene’ that stretch of DNA which can generate dozens of different proteins; on the other, a gene could be that unit which is translated into one specific polypeptide, i.e., we might call a ‘gene’ each individual spliced mRNA by assuming a ‘one mature mRNA-one protein’ relationship. Guimarães (1986), for instance, argued that gene identity could be univocal at the nucleic 17 Symptomatically, Black (2000) considers protein diversity resulting from alternative RNA splicing as a major challenge for bioinformatics and post-genome biology.
Chapter 2. The problem of the gene
43
acid level if the fully processed and mature RNA had a single function, an idea he himself rejected later (Pardini, Guimarães 1992), due to alternative translation modes. But, despite alternative translation, to treat mature mRNA as the gene has in itself a number of counter-intuitive consequences. It would mean, for instance, that genes exist in the zygote only as possibilities and don’t show the permanence and stability typically ascribed to the genetic material. Moreover, genes would not be found in chromosomes and, sometimes, not even in the nucleus (Keller 2000). Due to split genes and alternative splicing, the linear correspondence between a gene and a transcription unit does not hold, and model B fails too. A putative solution, then, is to move from model B to model C, treating units in the genome as smaller in size. Model C can be related to a controversy in the literature regarding the inclusion in or exclusion of introns from a gene. Some authors, as, for instance, Waters (1994), take a ‘it doesn’t matter’ view, in which introns may be either included in or excluded from genes. Others, such as Epp (1997), insist that introns should be excluded. Epp claims that an acceptable term already exists for the entire DNA region, including both exons and introns, namely, ‘transcription unit’, so that an ‘exons-only’ definition of a gene should be advocated. His proposal of a definition for ‘gene’ refers to the nucleotide sequence that stores ‘information’ specifying the order of monomers in a functional polypeptide or RNA, or in a set of closely related isoforms. In fact, Epp’s proposal corresponds to Fogle’s (1990) model C, which seems at first capable of assimilating alternative splicing, by treating exons as the structural units in the genome and, consequently, rescuing the idea that a gene is a unit by redefining genes as sets of exons sharing a common transcript.18 Could this be a putative solution to the gene problem? The answer seems to be ‘No’. Model C faces the problem that there are patterns of RNA splicing which result in transcripts which differ from one another by the presence or absence 18
This definition of gene was employed by Venter and colleagues (2001: 1317) in one of the papers which reported the draft sequence of the human genome: “A gene is a locus of cotranscribed exons. A single gene may give rise to multiple transcripts, and thus multiple distinct proteins with multiple functions, by means of alternative splicing and alternative transcription initiation and termination sites”. As we can see in this passage, they argue for this definition of gene because of the challenges to Fogle’s model B, but they don’t take into account the problems with model C.
44
Genes, Information, and Semiosis
of exons corresponding to trailer sequences (Henikoff et al. [1987] offers an example, discussed by Fogle 1990). As this model includes the exons corresponding to trailer sequences, this feature is enough to falsify it. Nevertheless, Model C can be easily saved, in principle, by a slight modification, which leads to model D, including only coding exons. In this case, any difference in the length of trailer sequences becomes irrelevant. But alternative splicing can also affect the size and coding region of exons, as shown by Schulz et al. (1986) study of Drosophila Eip 28/29 gene (see Fogle 1990). This gene is composed of four exons and produces two mature mRNAs which differ by 12 nucleotides in one of the exons. These 12 nucleotides are part of an intron in one of the transcripts, and part of a coding exon in the other one. Therefore, alternative splicing challenges also model D, and, in fact, have consequences which are far more radical than it might seem at first, as shuffling of exons to produce multiple messages is not the only mechanism involved in this phenomenon. The conclusion we reach, thus, is that none of the structural models discussed above holds. If we treat them as forming the whole set of possible structural models, as it seems plausible to do, we can see why the idea of the gene as a structural unit is in crisis. It is clear, however, that we can understand the situation in a different way, since alternative splicing affecting coding exons simply shows that model D is not absolutely general. But where in biology do we have entirely general models? Why should we demand such a generality from models of the structural gene? It seems clear that the most reasonable conclusion, as regards this latter model, is that it shouldn’t be discarded on the grounds of a particular kind of alternative splicing, since the model remains useful despite possible exceptions. However, we must then face additional challenges to the classical molecular gene concept, found in our current understanding of the structure and dynamics of genomes. For instance, mRNA edition (Hanson 1996; Lewin 2004) makes all the structural models examined by Fogle (1990) look problematic. In this process, individual bases are added to or deleted from mRNA during processing. As Lewin (2004: 742) puts it, “RNA editing is a process in which information changes at the level of mRNA. It is revealed by situations in which the coding sequence in an RNA differs from the sequence in DNA from which it is transcribed.” Therefore, in
Chapter 2. The problem of the gene
45
the case of proteins synthesized out of edited mRNAs, no corresponding sequences can be found on DNA even considering splicing. Furthermore, the nucleotide sequences found in the final version of mRNAs are not (at least not directly) transmitted to the next generation (see Keller 2005; Griffiths, Neumann-Held 1999). Another substantial difficulty results from the existence of alternative translation modes in genetic systems (Pardini, Guimarães 1992). Regulatory mechanisms can lead to the choice of alternative initiator and terminator codons when one mRNA is translated. Alternative initiations can lead both to partially homologous products — if they are in the same reading frame — or to non-homologous polypeptides — if the reading frames are different. In turn, alternative terminator codons can lead to polypeptides of different sizes. In these cases, there is no univocal correspondence between sequences of nucleotides in DNA and single products. Considering these additional challenges, it becomes clear that even model D is not in a comfortable situation as a basis for understanding genes as structural units in the genome. But there is another possibility to examine. After concluding that none of the structural models holds, Fogle (1990) discusses if it is possible to find a complexly organized unit representing a discrete region of DNA which is distinct from other also complexly organized units. But then he faces yet other challenges to the gene concept, such as gene overlapping and transsplicing. As the coding sequences of a gene are contained in only one of the two complementary strands of DNA, it is not unusual to find two different genes overlapping, each located in a different strand. Nevertheless, there are also genes overlapping in the same strand of DNA, as it is the case, for instance, of several genes in the rather compact genome of HIV.19 And, to consider an even more bewildering situation, we also find nested genes (i.e., genes inside genes) in the genome. It is not rare, for instance, that introns harbor inside themselves whole genes, as it is the case of intron 26 of the gene for neurofibromatosis type 1, which contains three genes nested inside it, each of them also containing introns. It can even be the case that nested genes are not, as it seemed at first, uncommon gene structures (Wang et al. 2000). And, moreover, while nested genes were until 19 It is important to stress, however, that gene overlapping are not restricted to compact genomes, such as those of viruses.
46
Genes, Information, and Semiosis
recently found to be usually around 1 kb long, Wang et al. (2000) discovered a nested gene in Drosophila which is 7.6 kb long. All these cases of overlapping pose obvious difficulties to the idea that genes are structural units in DNA. Differently from cis-splicing, the process by which introns are removed from a pre-mRNA, in trans-splicing, two RNA molecules encoded in different loci in the genome are spliced together. This process — which is defined by Hastings (2005: 240) as the “joining of two separate RNA molecules to form a chimeric molecule” — is found in a variety of organisms. We will focus our arguments here on a particular form of trans-splicing, called ‘spliced-leader (SL) transsplicing’. Parasitic trypanosomes offer a standard example of SL trans-splicing. They produce abundant amounts of a single 140nucleotide leader RNA, from which a 39-nucleotide portion, termed a ‘mini-exon’, is spliced to the 5’ end of protein-coding exons in primary transcripts. The 5’ mini-exon is present in all trypanosome mRNAs and is thought to assist in initiation of translation. Due to SL trans-splicing, diverse mRNAs — ranging from a minority to even 100% of the mRNA population in different organisms — exhibit a common 5’-sequence (Hastings 2005). SL trans-splicing shows a patchy phylogenetic distribution — the organisms showing SL transsplicing include euglenoids, nematodes (as C. elegans), flatworms (as Schistosoma mansoni), and even chordates (as the ascidian Clona intestinalis. See Vandenberghe et al. 2001) — and its evolutionary origins are still enigmatic (Nilsen 2001; Hastings 2005). Anyway, trans-splicing raises another difficult to the ‘gene as a unit’ concept (Fogle 1990): is the trans-spliced RNA the product of the merging of two units? Or is it itself a coding unit, despite its split origin in DNA? In all models examined above, a shared feature was the contiguity of a primary transcript located at one locus. In trans-splicing, even the structural integrity of the primary transcript is lost. But things can be even more puzzling. Consider, for instance, the S12 ribosomal gene encoded by chloroplast DNA, which appears to be organized in three exons which are transcribed separately, two on one DNA strand, and the third on the opposite strand (Koller et al. 1987; Fogle 1990). In this case, a contiguous message for translation is in fact built through trans-splicing. Another challenge to the classical molecular gene concept comes from genomic rearrangements through which genes are formed during
Chapter 2. The problem of the gene
47
ontogenetic processes (Pardini, Guimarães 1992). In the immune system, for instance, sequences coding for antibodies and T-cell receptors are assembled, during ontogeny, from segments which are far apart in germ cells. In this case, we observe many-to-one relationships between DNA sequences and RNA/polypeptide sequences, posing another difficulty to the unit concept. Generally speaking, systems which depend on high levels of variability to function properly, such as the immune and the nervous systems, show intense processes of permutation of modules and transcriptional motives during ontogeny. Moreover, an increase in the complexity of regulatory mechanisms underlying these processes, rather than in the number of genes, is responsible for huge differences between species in the case of these systems. Enard and colleagues (2002), for instance, argue that the human brain shows a higher molecular complexity than the brains of other primates, such as chimpanzees, orangutans, and rhesus monkeys, due to changes in gene expression patterns during recent human evolution. One of the most recent difficulties for the gene concept comes from the discovery of micro-RNAs with remarkable regulatory powers, coded by DNA sequences scattered throughout the genome, mostly in regions previously named ‘junk DNA’ (Fire 1999; Grosshans, Slack 2002; Hannon 2002; Lenz 2005). The problem resulting from this discovery is not new: some definitions of ‘gene’ refer only to protein-coding sequences, while others also include non-proteincoding regions. Therefore, according to some definitions, the sequences coding for micro-RNAs would count as genes, according to others, not. What is dramatic about the problem is its dimension: 98.5% of human DNA, for instance, corresponds to non-proteincoding sequences, much of it coding for RNAs with regulatory functions (Keller 2005). Venter et al. (2001) report that only 1.1% of the human genome corresponds to exons, while 24% are introns and 75%, intergenic DNA. The number of genes identified or estimated in the drafts of the human genome sequence published in 2001 came as a surprise both to the scientific literature and to the media (Leite 2005), and indeed added to the problem of the gene. Lander et al. (2001) reported 30.000-40.000, while Venter et al. (2001) identified 26.000-38.000 transcriptional units, numbers much smaller than the previous
48
Genes, Information, and Semiosis
estimations, ranging from 50.000 to more than 140.000 genes.20 These estimations were based on the interpretation of genes as structural and functional units, as if there were one gene for each protein believed to exist in the molecular repertoire of the human species (Leite 2005). The number of genes identified in the drafts of the human genome clearly shows that this is not the case, as Venter admitted in a press release sent by Science to journalists:21 The small number of genes […] supports the notion that we are not hard wired. We now know that the notion that one gene leads to one protein and perhaps to one disease is false. One gene leads to many different products and these products — proteins — can change dramatically after they are produced. We know that regions of the genome that are not genes may be the key to the complexity we see in humans. We now know that the environment acting on these biological steps may be key in making us what we are. (Venter 2001, cited by Leite 2005: 132)
Alternative splicing, which is a rather common phenomenon in mammalian genomes, is partly responsible for the economy of genes in the human genome. Recent genome-wide analyses indicate that 35– 59% of human genes produce alternatively spliced forms (Modrek, Lee 2002). Even though a significant portion of the predicted splicing variants are not functional, resulting from aberrant rather than regulated splicing, and, therefore, the frequencies of alternatively spliced gene products mentioned above are probably overestimated (Sorek et al. 2004), it is still the case that alternative splicing should be regarded as one of the most significant components of the functional complexity of the genome of our and other species (Modrek, Lee 2002). Both many-to-one and one-to-many relationships between DNA sequences and RNA/polypeptide sequences give support to a picture of molecular complexity in which the amount of genes is not the 20
But notice that there is a great degree of uncertainty and controversy about the estimation of the number of human genes. Wright et al. (2001), for instance, report a significantly higher number, 65,000–75,000 transcriptional units, with exonic sequences comprising 4%. They ground their estimation on the argument that only 1 out of 5 genes are reported by both Venter et al. (2001) and Lander et al. (2001), and also on the assumption that computational tools are not sufficient to identify all transcriptional units. 21 Notice also the denial of deterministic theses about the human nature which have been previously so common in the rhetoric of genome projects.
Chapter 2. The problem of the gene
49
crucial feature, but rather complex information networks, such as those mediated by transcription factors (see, e. g., Lee et al. 2002), and intricate patterns of gene expression, which allow for a huge diversity of proteins and RNAs based on a limited number of genes (see, e. g., Szathmáry et al. 2001; Maniatis, Tasic 2002).
7. How the gene concept became a problem: why is the gene not a functional unit? In view of the difficulties faced by the idea that genes are structural units, we should investigate the alternative of treating them as functional units. If one wants to understand gene function, it is necessary to examine the nature of gene expression, since it is by being expressed that a gene can have significance to the cell. Nevertheless, gene expression shows that the idea of the gene as a functional unit also faces important difficulties. The classical model of the gene as a unit of function (Benzer’s [1957] cistron) is grounded on the idea that a gene produces a single polypeptide, which, in turn, has a singular function. But the complexity of gene action in the cellular context makes it quite difficult to maintain the idea of a unitary relationship between a gene and its function. The context-dependence of gene action clearly shows that it makes no sense to ascribe a single function directly to a DNA locus, without taking into account in which context that locus is expressed. One manner of emphasizing the context-dependence of gene function is to properly consider the role of regulation in living systems. Differences in animal designs and complexity, for instance, are mostly related to changes in the temporal and spatial regulation of patterns of gene expression (Carroll et al. 2005), and not so much to the evolution of genes themselves, as shown by sequence comparison between several animal genomes. Peltonen and McKusick (2001) argue that a shift from a focus on gene action to an emphasis on gene regulation is one of the features of a possible paradigm change in current molecular biology and genetics. Keller (2000), in turn, perceives the seeds of this change emerging a long time ago, when Jacob and Monod’s model of gene regulation brought into focus the idea of gene activation rather than action. Regulation is a process which entails an influence of higher-level processes on molecular
50
Genes, Information, and Semiosis
processes, such as transcription, RNA splicing, translation etc., i.e., it involves a kind of process which has not been clearly conceptualized yet in biological thought, namely, downward determination (see Ch. 7). The time and place in which a given set of genes is activated or not crucially depends on downward regulation, and this regulation is something to which genes are submitted, not something that genes do, command, control, program, etc. Regulation highlights a claim we put forward in Ch. 1: it is not DNA that does things to the cell; rather, it is the cell which does things with DNA. Furthermore, regulation brings with it a demand for semiotic analysis, since signaling is a key feature of each and every regulatory pathway in living systems. Consider, moreover, that many genes, if not most, show pleiotropic effects, affecting a variety of traits. We can highlight, then, an important deleterious effect of the unit conception, particularly when we consider the discourse about genes in society, and the ideas about genes and their functions that appear in policy documents concerning topics such as transgenic organisms. A number of effects of a pleiotropic gene are often taken to be secondary — or even non-existent — and the gene is treated as an entity producing one primary genetic effect, related to a trait of interest. Thus, if one is discussing a genetic modification in maize which makes this crop more productive, it can be the case that one neglects many other effects that such a modification may have, some of them consequential to human health (think, say, of the possibility that the gene at stake has effects on other genes that may produce an allergenic compound) or to the environment (consider a gene that may have as a secondary outcome a harmful effect on a species of insect, which happens to be important to the pollination of a number of plants). This stresses one of the major problems with the model of the gene as a functional unit: it is only possible to hold a one-to-one relationship between gene and phenotype if a narrowly defined context is proposed, and this in turn can be only made by denying that other contexts exist. Needless to say, this is not a very smart approach when one considers how genes can act in the very complex contexts one finds in organisms and environments. Nevertheless, the unit concept gives support to such undesirable simplifications of the relationships between genes, organisms, and environments. Even if we consider a single protein coded by a gene, it will be difficult to sustain the idea of a functional unit, since many proteins
Chapter 2. The problem of the gene
51
are multifunctional. Among many possible examples, we can mention the enzyme tryptophan synthetase, which has two catalytic functions: while its α subunit catalyzes the conversion of 1-(indol-3-yl)glycerol 3-phosphate to indole and glyceraldehyde 3-phosphate, its β subunit catalyzes the condensation of indole and serine to form tryptophan. Things become even more problematic when we consider alternative splicing, by means of which a DNA locus can code for multiple polypeptides. All these arguments go against the idea that the criterion for the material gene might be that the gene is a unit of function.
8. What is a gene, after all? Our current knowledge about the physical organization and dynamics of genomes brings to collapse the delicate juxtaposition of the molecular and the Mendelian gene established in the classical molecular concept. The advances discussed in the previous sections resulted in an overwhelming proliferation of instrumentally formulated genetic entities, an expanding zoo of heterodox genetic entities (Falk 1986: 164). Historically, it became evident that genes are neither discrete (there are overlapping and nested genes), nor continuous (there are introns within genes); they do not necessarily have a constant location (there are transposons), and they are neither units of function (there are alternatively spliced genes and genes coding for multifunctional proteins, and gene action is strongly dependent on cellular and supracellular contexts), nor units of structure (there are many kinds of cisacting sequences influencing transcription, split genes, etc.) (cf. Falk 1986; Fogle 1990). When there are so many problems with the properties used to define a concept, it is natural to ask what, after all, is the entity which is being defined. Fogle (1990, 2000) vigorously argued against the maintenance of the gene-as-a-unit concept — regardless of whether as a unit of inheritance, structure, function and/or information. It is this concept that cannot be reconciled with our current knowledge about the structure and functioning of genomes. This opens the door for an attempt to save the gene concept, rather than abandoning it, as Keller (2000) proposed (but cf. Keller 2005, see Ch. 1): the gene can be redefined in such a manner that the unit concept is dispensed with. To this effect, Fogle (1990) offers an interesting alternative interpretation of genes as
52
Genes, Information, and Semiosis
sets of domains in DNA and Pardini and Guimarães (1992) propose a ‘systemic concept of the gene’. We will come back to these proposals later. This is no minor change in our view about genes. After all, the main historical baggage of this concept lies precisely in the understanding of genes as basic units of life (Keller 2005), which in fact predates the gene concept itself (Fogle 2000). It is true that there is no uniform agreement among scientists and philosophers as to the conceptual incompatibility of the Mendelian and molecular pictures of the gene (cf. Moyle 2002). Moyle mentions, for instance, Waters’ (1994) claim that there is no tension between classical and molecular conceptions of the gene, as the traditional concept can be equated with the mRNA coding region of DNA (i.e., a cistron). We will critically address Waters’ view below, but, for the moment being, we can point out that many of the above criticisms to models of the structural gene show how difficult it is to keep the classical idea of the gene as a unit by merely relying on DNA coding regions. Merely appealing to the concept of cistron does not offer an easy way out of the problem of the gene, as Moyle (2002) suggests. But, even if we agreed, for the sake of the argument, with Moyle’s claim, it would still be the case, as he himself argues, that this compatibility “… is often not as comfortable as many portray” (Moyle 2002: 723).
9. The gene in the era of genomics and systems biology Recent advances in molecular biology, genomics and proteomics made it more and more difficult to conceive of genes as units. It is now quite clear that biological information operates at multiple hierarchical levels, in which complex networks of interactions between components are the rule, and, consequently, the understanding of the dynamics and even the structure of genes demands that they are located in complex informational networks and pathways (Ideker et al. 2001). These features of genomic systems have pushed researchers into adopting a ‘systemic’ perspective, which has given rise, in turn, to the current wave of ‘systems biology’ in the fields of molecular biology, genomics, and proteomics.22 22
It is important to notice that not all researchers in these fields are enthusiastic about the surge of ‘systems biology’. For instance, Claverie (2001) critically
Chapter 2. The problem of the gene
53
Systems biology is often presented as a non-reductionistic approach (Chong, Ray 2002; Kitano 2002; McCarthy 2004; Stephanopoulos et al. 2004; Barabási, Oltvai 2004; Nature, 2005). Many genomic researchers seem quite eager, indeed, to declare that they have overcome ‘fallacies’ such as determinism and reductionism (see, e. g., Venter et al. 2001: 1348), even though a sort of embarrassed determinism (cf. Leite 2005) lives on in their writings. It is not clear at all, at present, what ‘systems biology’ really means in these fields (Keller 2005), and, furthermore, it can be put into question if it is really such a non-reductionistic approach as many of its advocates claim (Bruni 2003; Morange 2006), much in the same sense as systems ecology was previously charged of being nothing but a largescale reductionistic approach (e. g., Levins, Lewontin 1980; Bergandi 1995).23 When one reads, for instance, that “a key aim of postgenomic biomedical research is to systematically catalogue all molecules and their interactions within a living cell” (Barabási, Oltvai 2004: 101), one might suspect that what is at stake in this approach, after all, is not so much a non-reductionistic, but rather a large-scale reductionistic approach. And, when we read, in the follow-up of the same argument, that “rapid advances in network biology indicate that cellular networks are governed by universal laws” (ibid. id.), a sense of bewilderment is inevitable: if cellular networks are indeed governed by universal laws, why would it be necessary to catalogue all molecules and interactions? Might it be the case that other, more systemic approaches would appraises the arguments of ‘an increasing number of researchers’ who have been saying that “the old and classical ‘reductionist’ approach would be totally inadequate to figure out the function of all genes” (Claverie 2001: 1257). In contrast with their proposal (which is, in his view, “reminiscent of the old general system theory”) that “… complex genetic networks should be studied as a whole, using ‘new theoretical approaches’, according to the premise that nondeterministic and/or chaotic phenomena might govern the functioning of the human genome” (Claverie 2001: 1257), he favors “… the use of simple hierarchical regulatory models in conjunction with the spectacular development of high-throughput analyses (microarray, two-hybrid system, proteomics, chemical screening, etc.)” (Claverie 2001: 1257), which will be sufficient, in his view, to generate most of the significant results in functional genomics. But Claverie is an isolated voice, since most researchers in these fields feel that there are problems with the reductionist research strategy (Leite 2005), even though it is not clear that the alternative they are seeking is not reductionist too. 23 In Chapter 1, we also argued that there problems with the treatment of ‘biological information’ in ‘systems biology’ (see also Bruni 2003).
54
Genes, Information, and Semiosis
reveal in a more efficient (and cheaper) way how cellular networks function? Despite the doubts expressed above, it is clear that there is at least a tension between previous, ‘more reductionistic’ approaches in molecular biology, and the current tendency to raise the level of analysis of cell systems back to the level at which cellular phenomena indeed emerge. Even though it is not clear that the current move in molecular biology and genomics can be indeed treated as a shift from ‘reductionism’ to ‘whole-istic’ biology (Chong, Ray 2002), it seems to be indisputable that a movement towards ‘less reductionistic’ approaches is really taking place in these fields. Rather than being the climax of reductionistic research in biology, the genome projects seemed to have shown the limits of reductionism in biological research, as it was observed that it is not possible to simply derive function from structural information about nucleotide sequences (Keller 2000).24 The transition from structural to functional genomics may indeed show a new starting-point to molecular biology, inspired by less reductionistic or, maybe even, non-reductionistic grounds. We find in the literature, for instance, a clear recognition that by merely contemplating DNA sequences, one cannot really predict the precise functions of coding regions (Stephens 1998: R47), and not even define all genes and their transcripts (Camargo et al. 2001). Furthermore, we notice approaches to the problem of gene function which seem to diverge, at least to some extent, from the reductionistic tendency that dominated biology in the second half of the 20th century. Hieter and Boguski (1997: 601), for instance, define functional genomics as “the development and application of global (genomewide or system-wide) experimental approaches to assess gene function by making use of the information and reagents provided by structural genomics”. In an explanation about a new project from the US Department of Energy (where the human genome initiative was born) called “Bringing Genomes to Life Program”, we also find: 24
Leite (2005) discusses other interesting aspects of genomics, such as the tension between the few results in new approaches to human health and modest economic success, on the one hand, and the hyperbolic promises for the future in several papers and books in the field, on the other. He also offers an interesting analysis of the deterministic/reductionistic and even eugenicist rationale and rethorics of genome projects, especially in the editions of Nature and Science in which the draft sequences of the human genome were published, in 2001, and in popular science books, such as Watson (2000) and Watson, Berry (2003).
Chapter 2. The problem of the gene
55
The current paradigm in biology — variously described as ‘single gene’, ‘reductionist’, or ‘linear’ — is not likely to be successful on its own. […]. Knowing the function of all genes in the genome, by itself, will not lead to understanding the processes of a living organism. (cited by Keller 2005: 4)
In his opening address at the first annual conference of MIT’s Computational and Systems Biology Initiative, then president of MIT Charles M. Vest said: Until now, biologists have learned more and more about the detailed structure and functions of the molecular components of life, but we have not yet understood how individual components are networked to control physiology. [...] Now we are in a position to begin the search to understand our molecular machines and cell circuitry — how the parts are connected and how they operate. In a third revolutionary transformation, biology will become a systems science. (Vest 2003, cited by Keller 2005: 5)
In sum, if it is not clear that these current changes in biological research are indeed leading us to non-reductionistic approaches to living systems, it seems to be the case, at least, that the reductionistic tendency which has been typical of molecular biology for so many years is under tension now. Our current understanding of genomes indicates that we should move beyond the treatment of genes as units of structure and function which, secondarily, interact in complex networks. In contrast to beanbag and deterministic views, genes themselves should be thought of in a systemic manner/context, as emergent structures produced by the network of interactions into which stretches of DNA are embedded. This idea hints at a process-interpretation of genes, to which we will return later. These changes in our understanding of genes have also important empirical consequences. For instance, the complex dynamics of gene structure and function highlights the limits of gene annotation, through which researchers try to locate genes in sequence data and associate them with specific functions. Annotation is currently not enough to detect genes, depending on validation by experimental approaches, and our growing understanding of genomes seems to point out that the situation is likely to remain the same. Tupler and colleagues, for instance, describe the limits of the search for genes in sequence data as follows:
56
Genes, Information, and Semiosis [...] the existence of a related gene sequence does not mean that there is a corresponding protein: the sequence could be a non-expressed pseudogene. Even if a related sequence is an expressed gene, we do not know whether the two related genes are simultaneously expressed in the same cell, or are differentially expressed — for example, in a tissue — or development-specific manner. Expression studies will be required to complement genomic information. […] the full value of the genomic information can be realized only when it is coupled with appropriate biochemical studies. (Tupler et al. 2001: 833)
Symptomatically, doubts about the status of the gene concept are found today not only in philosophical but also in empirical papers, in a manner which is suggestive — if we adopt a Kuhnian perspective — of a crisis in the paradigm that dominated molecular biology since the proposal of the double helix model.25 It is interesting to see empirical scientists, usually not so keen of conceptual discussions, expressing deep concerns about the ways of using the term ‘gene’ or even its utility itself. We will mention here just some recent examples. Wang et al. (2000), in a study of the origin of a particular gene and the complex modular structure of its parental gene, claims that this structure “[...] 25 Within genetics and molecular biology, there is indeed a growing feeling of a change of paradigms. Strohman (2002: 703), for instance, argues that ‘systems biology’ can be seen “in terms of Kuhn’s observation concerning similar junctures of crisis and revolutions in the history of the physical sciences”. In his view, a transition from ‘normal’ to ‘extraordinary science’ can be taking place under our noses, pointing out to a revolutionary change in our understanding of the genotype-phenotype relationship. This may bring about a ‘rediscovery’ of the complexity of this relation, eventually leading to the replacement of the dominant gene-centric approach, in which the dynamic protein networks that generate phenotype from genotype have been largely ignored, by a new account, in which genes and dynamic systems will have equal standing. Peltonen and McKusick (2001: 1226), in turn, describes a series of conceptual shifts (reflected in changes in research targets, methods, and strategies) which they see as symptoms of a paradigm change: e. g., structural genomics → functional genomics; genomics → proteomics; monogenic disorders → multifactorial disorders; specific DNA diagnosis → monitoring of susceptibility; analysis of one gene → analysis of multiple genes in gene families, pathways, or systems; gene action → gene regulation; etiology (specific mutation) → pathogenesis (mechanism). Even though we can see in this catalog of changes a non-reductionistic tendency, it is not clear, as we argued above, that this is indeed the direction of these shifts. Further epistemological research is needed to come to grips with the current shifts in the epistemic bases of molecular biological research.
Chapter 2. The problem of the gene
57
manifests the complexity of the gene concept, which should be considered in genomic research” (Wang et al. 2000: 1294), for instance, when one tries to predict a gene from genome data (Wang et al. 2000: 1300). Recognizing the problems faced by the gene concept, they also consider that their finding [...] may add to the classical concept of genes, which has been modified with the discoveries of operons, introns, overlapping genes, alternative splicing, multiple polyadenylation sites, complex promoters, and nested genes. (Wang et al. 2000: 1300)
Kampa and colleagues, for instance, argue that their observation that 49% of the transcribed nucleotides in human chromosomes 21 and 22 amount to novel classes of RNA transcripts, while only 31,4% correspond to well-characterized genes, “[...] strongly support the argument for a re-evaluation of the total number of human genes and an alternative term for ‘gene’ to encompass these growing, novel classes of RNA transcripts in the human genome” (Kampa et al. 2004: 331, emphasis added). They do not suggest that we should abandon the term ‘gene’ altogether, but observe that [...] the use of the term ‘gene’ to identify all the transcribed units in the genome may need reconsideration, given the fact that this is a term that was coined to denote a genetic concept and not necessarily a physical and measurable entity. With respect to the efforts to enumerate all functional transcribed units, it may be helpful to consider using the term ‘transcript(s)’ in place of gene. (Kampa et al. 2004: 341)26
Finally, the papers reporting the draft sequence of the human genome in 2001 also recognize the complexities surrounding genes and their functions. Lander and colleagues (2001), for instance, offer an inventory of several genomic elements which do not fit into the picture of a simplistic chain from genes in DNA to a single protein to a particular trait, and also emphasize that the system which regulates gene expression is as intricate as it is far from being understood. Moreover, it is also important to stress that the gene-centric and 26 Incidentally, it is interesting to note that Kampa and colleagues’ remark that the term ‘gene’ was coined to ‘denote a concept’, not necessarily a ‘physical and measurable entity’, lends support to Falk’s (1986) claim that an instrumentalist view about the gene became once again quite frequent among practicing scientists in genetics and molecular biology (see Chapter 1 and below).
58
Genes, Information, and Semiosis
deterministic views which characterized for so long the genome projects, although still present, appeared in the special numbers of Nature and Science in which those reports were published in a somewhat mitigated manner. This was clearly a result from the wave of both criticisms and data which have shown in the 1990s how difficult it is to understand the genome in accordance with such views.27
10. Some reactions to the problem of the gene Conceptual variation and ambiguities in the gene concept A number of reactions to the crisis of the gene concept can be found in the literature, as mentioned in Ch. 1, ranging from the idea of leaving it aside and trying to build new conceptual tools for understanding how the genetic material is organized and functions (Keller 2000. But cf. Keller 2005) to proposals of manners of saving it (Falk 2001; Hall 2001). A major problem faced by the gene concept is proliferation of meanings (Fogle 1990, 2000; Moss 2001). Conceptual variation and ambiguities have been a feature of the gene concept throughout its whole history28, and a number of authors consider that they even have been heuristically useful in the past (Kitcher 1982; Burian 1985; Falk 1986; Griffiths, Neumann-Held 1999; Stotz et al. 2004). Griffiths and Neumann-Held (1999), for instance, stress that a single definition for genes may not be necessary or even desirable, since different gene concepts may be useful in different areas of biology. Kitcher (1982) 27
For a discussion about this issue, see Leite (2005). Even before the wealth of data showing that the genetic material is much more complex than it first seemed, Carlson (1966: 259), in his The Gene: A Critical History, offered an impressive list of different interpretations about what is a gene: “The gene has been considered to be an undefined unit, a unit-character, a unit factor, a factor, an abstract point on a recombination map, a three-dimensional segment of an anaphase chromosome, a linear segment of an interphase chromosome, a sac of genomeres, a series of linear sub-genes, a spherical unit defined by target theory, a dynamic functional quantity of one specific unit, a pseudo-allele, a specific chromosome segment subject to position effect, a rearrangement within a continuous chromosome molecule, a cistron within which fine structure can be demonstrated, and a linear segment of nucleic acid specifying a structural or regulatory product”. 28
Chapter 2. The problem of the gene
59
states that we can allow the term ‘gene’ to have a heterogeneous reference potential, since different gene concepts may be useful for different research goals. The recognition of the heuristic role of conceptual variation does not preclude, however, a worry about the possibility that it can also lead to important difficulties. Falk, for instance, considers that the pluralism found in the current picture about genes [...] brought us […] dangerously near to misconceptions and misunderstandings […]. It is to be hoped that a more reflective mood among the investigators themselves about the meaning of their concepts may be enough to avert the dangers and so will allow for a continued reaping of the fruits of success. (Falk 1986: 173)
Fogle argues that, despite proposed methodological advantages for the juxtaposition of ‘gene’ concepts it is also true […] that confusion and ontological consequences follow when the classical intention for ‘gene’ conjoins a molecular ‘gene’ with fluid meaning. (Fogle 1990: 350)
Keller (2005) argues that many problems arise from ambiguities in the usage of the term ‘gene’, particularly calling attention to the difficulties regarding gene counting, since the value obtained will vary by 2, 3 or more orders of magnitude depending on how genes are defined, and it is not always evident what one is counting (see also Keller 2000). This conceptual variation arguably results from a change in our attitude towards the gene as a theoretical concept. Falk (1986) argues that the difficulties faced by the gene concept eventually led us back to an instrumentalist view: “Today the gene is not the material unit or the instrumental unit of inheritance, but rather a unit, a segment that corresponds to a unit-function as defined by the individual experimentalist’s needs” (Falk 1986: 169, emphasis in the original). ‘Gene’ sometimes means a specific unit in the genotype, sometimes it is a collective term for genetic units and often is avoided completely (Falk 1986: 165). Terms like ‘cistron’, ‘copies’, ‘genetic element’, etc. are often used to refer to DNA sequences involved in ‘gene’ action. Sometimes the term ‘gene’ is applied only to the coding region, sometimes it refers to coding and regulating sequences alike (Falk 1986: 166), and so on.
60
Genes, Information, and Semiosis
Fogle (1990) argues that his analysis of the situation of the geneas-a-unit concept supports Falk’s (1986) claim that the gene is currently seen once again as an instrumentalist, pragmatically flexible construct, which can be adjusted to the diverse needs of researchers in different fields. But he offers a mostly negative appraisal of this state of affairs, considering that retreating to the view that the material gene is an instrumental construct confounds meaning and hinders specification of gene properties. In a similar vein, Rheinberger (2000) treats genes as ‘epistemic objects’ in molecular biology, i.e., as entities whose name is introduced as a target of research rather than to designate something with which researchers are acquainted. The set of experimental practices used by scientists plays a fundamental role in the constitution of their understanding of an epistemic object. This helps, in turn, explaining conceptual variation: differences in experimental practices used by diverse communities of scientists lead to differences in the meaning and application of ‘gene’, as an epistemic object (Stotz et al. 2004).
Searching for a single, inclusive gene concept Conceptual variation is particularly distressing for those who regret that the term ‘gene’ has no “precise universally accepted molecular definition” (Epp 1997: 537). For them, attempts to reduce the diversity of definitions of ‘gene’ to a single gene concept are particularly welcome. One such attempt is found in Waters’ proposal: “[The] fundamental concept […] is that of a gene for a linear sequence in a product at some stage of genetic expression” (Waters 1994: 178). By using a number of open clauses, he intends to accommodate the challenges to the gene concept. For instance, if one asks whether introns should be included or not in genes, Waters can answer that it depends on which particular ‘linear sequence in a product’ at which ‘stage of genetic expression’ one is addressing. If we focus on the process of transcription at the stage of pre-mRNA, then introns should be included in genes. But, if we focus on the polypeptide chain, introns should not be included. Waters’ attempt follows from his argument that, while discussing the gene concept, philosophers have often conflated ambiguities of
Chapter 2. The problem of the gene
61
application and conceptual ambiguities (Waters 2000). It is from this appraisal that he launches a criticism of the idea that there is no such thing as a gene at the molecular level, but just promoters, operators, introns, exons, etc. We will come back to this idea below, when discussing Fogle’s proposal that genes can be conceived as sets of domains in DNA. Waters, in contrast, supports the idea that there is a clear and uniform way of understanding genes at the molecular level, allegedly grasped by the gene concept he proposes. But it is not clear at all that Waters’ proposal helps solve the problem of the gene. Even if his definition reflects the current usage of the term ‘gene’, it is doubtful whether it can help clarify the conceptual issues raised by the growing understanding of the complexity of gene expression (Griffiths and Neumann-Held 1999). If his definition is accepted, then several ‘genes’ will come into being at different stages of the expression process. The entire DNA sequence from which a pre-mRNA is transcribed can be treated, for example, as a gene. But those parts of the sequence corresponding to mature mRNA, i.e., exons, can be treated as genes in an equally legitimate manner. If a variety of mature mRNAs is produced by means of alternative splicing, then several ‘genes’ would stem from one single ‘gene’. As Griffiths and Neumann-Held (ibid., id.) sums up, “… it seems strange, to say the least, to assert that single genes on the DNA correspond to a variety of genes on the mRNA”. It is not clear in what sense Waters’ proposal would be explanatorily more powerful than the set of terms currently employed to describe empirical data in molecular biology, such as ‘noncoding regions’, ‘pre-mRNA’, ‘mature mRNA’, ‘intron’, ‘exon’, etc. There seems to be no good reason for defining genes in such a way that they end up being conflated with these established terms in the field.
The process molecular gene concept We agree with Kitcher (1982) and Griffiths, Neumann-Held (1999) that it is neither necessary nor desirable to have a single definition for ‘gene’; rather, we need different gene concepts, useful in different areas of biology, with different theoretical commitments and research practices. Nevertheless, even for those who think of conceptual variation as a desirable feature in our understanding of genes, it is the
62
Genes, Information, and Semiosis
case that we should clearly distinguish between different concepts and their domains of application (e. g., Falk 1986; Griffiths, NeumannHeld 1999). Griffiths and Neumann-Held (1999) attempt to organize the variety of gene concepts by proposing a distinction between the ‘molecular’ and the ‘evolutionary’ gene. An ‘evolutionary gene’, as introduced by Williams (1966) and elaborated by Dawkins (1982, 1989), amounts to “any stretch of DNA, beginning and ending at arbitrarily chosen points on the chromosome” that can be treated as “… competing with alelomorphic stretches for the region of chromosome concerned” (Dawkins 1982: 87). The theoretical independence of evolutionary genes in relation to molecular genes comes at a price, as one cannot defend assumptions about the former by appealing to the validity of these assumptions for the latter. Particularly, evolutionary genes cannot be identified with specific stretches of DNA (contra Dawkins). Nevertheless, the evolutionary gene concept has been introduced precisely on the grounds of our knowledge about molecular genes. If Griffiths and Neumann-Held are right — as we think they are —, this is not a legitimate move, and claims about evolutionary genes should be supported in the context of evolutionary, rather than molecular, biology (see also Sterelny, Griffiths 1999).29 They introduce an interesting proposal to this effect, namely, that evolutionary genes are best thought of as units of particulate inheritance, which was one of the first meanings ascribed to the term ‘gene’. An evolutionary gene would be, then, a theoretical entity, conceived as “a heritable potential for a phenotypic (or extended phenotypic) trait” (Griffiths, NeumannHeld 1999: 661). This definition has the interesting feature that it sets the evolutionary gene free from theoretical constraints from molecular biology, by avoiding any identification of evolutionary genes to particular DNA sequences. This may seem odd at first, since the idea that genes, no matter how we understand them, are segments of DNA 29 It is important to highlight that the conclusion that it is not legitimate to ground evolutionary genes on our knowledge about molecular genes has important consequences to the prospects of gene selectionism, since it is not the case that we can assume the latter as the right theoretical perspective in evolutionary biology on the grounds of our knowledge about molecular genes. The evolutionary gene concept has to be defended on its own ground, and the same is true of gene selectionism as a whole. Griffiths, Neumann-Held (1999) and Sterelny, Griffiths (1999) discuss other problems in Dawkins’ treatment of evolutionary genes, but we will not deal with these other criticisms here.
Chapter 2. The problem of the gene
63
is deeply entrenched. But, as Griffiths and Neumann-Held’s molecular process gene concept suggests, it may be a legitimate move. In turn, the molecular gene is, roughly speaking, a DNA sequence that codes for a polypeptide or RNA. This concept raises a multitude of problems, which point to a tension between two theoretical goals: on the one hand, to identify genes with particular segments on chromosomes (Kitcher’s [1982] ‘segmentation problem’, see above); on the other hand, to make genes central elements in the developmental explanation of phenotypic traits. We might keep the idea that a gene is a linear DNA sequence but abandon the idea that it has a single developmental role, defining it, for instance, as “a DNA sequence corresponding to a single ‘norm of reaction’ of genes products across various cellular conditions” (Griffiths, Neumann-Held 1999: 658). In this approach, the unit of development corresponding to each gene would become a disjunction of possible consequences under a variety of epigenetic conditions. Griffiths and Neumann-Held (1999) call this ‘most conservative response’ to the problem of the gene ‘the contemporary molecular gene concept’. They claim, however, that we should abandon the goal of identifying genes with particular segments on chromosomes in favor of the second aim, to understand genes as developmentally meaningful units. This movement leads to the ‘process molecular gene concept’ (Griffiths, Neumann-Held 1999; Neumann-Held 2001), in which genes are not treated as ‘bare DNA’, but rather as the whole molecular process underlying the capacity to express a particular product (a polypeptide or RNA). They characterize this concept as follows: “… ‘gene’ denotes the recurring process that leads to the temporally and spatially regulated expression of a particular polypeptide product” (Griffiths, Neumann-Held 1999: 659). This alternative builds the different epigenetic conditions which can affect gene expression into the gene. To understand this move, we can consider Epp’s (1997) proposal that we need to separate two distinct concepts related to genes, namely, a specification of ‘what is a gene’, and an indication of ‘how it is used’. But, while Epp claims that the term ‘gene’ points to the former, Griffiths and Neumann-Held argue that it rather points to the latter. They take, thus, a more processthan substance-oriented view on genes, stressing how they are used, rather than what they are as physical entities. They justify this option
64
Genes, Information, and Semiosis
by emphasizing that “the concept of the gene has always been intimately linked to how genes are used in development” (Griffiths, Neumann-Held 1999: 658). Some problems faced by the ‘gene’ concept also support their move. Consider, for instance, the case of regions in overlapping genes which are inside the open reading frame of a gene while function as promoters for other overlapped genes. Such situations suggest that whether a region in DNA is part of a particular gene or not depends on which function it performs relatively to that gene: [...] it is the role of a particular DNA sequence in a developmental system […] that influences whether the sequence is used (and can be described) as an intron or a coding region, as well as whether it is seen as a promoter or as part of an open reading frame. (Griffiths, Neumann-Held 1999: 659)
Therefore, functional descriptions of regions in DNA, such as ‘gene’, ‘promoter’, ‘enhancer’, cannot be explained merely in terms of structural descriptions, such as those of specific sequences in DNA. A structural description is, at best, a necessary condition for the functional description to apply, but not a sufficient condition, given the context-dependence of the function a given DNA region performs. Moreover, the process nature of the concept arguably makes it possible to accommodate anomalies which the classical molecular or, for that matter, the contemporary molecular gene concept has difficulty in facing, such as alternative splicing or mRNA editing. The key for dealing with these anomalies is the fact that the molecular process gene concept includes the particular processes involved both in alternative splicing and mRNA editing. Griffiths and Neumann-Held (1999) also offer a commentary on important similarities between the gene concept they propose and the classical molecular concept: first, the gene’s function of coding for a polypeptide (or RNA) is preserved; second, the gene still includes specific segments of DNA. The major difference, however, is that the gene is not identified with stretches of DNA alone, but rather with the processes in whose context these sequences take on a definite meaning.30 30
In this book, we will consider two alternative approaches to the consequences of our semiotic analysis, a more conservative and a more radical. In the latter, genes are treated as processes, as in Griffiths and Neumann-Held’s ‘molecular process gene concept’. The more conservative interpretation was reported in a
Chapter 2. The problem of the gene
65
It is also consequential that a process approach to the gene is not exclusively found among philosophers, but also among scientists, as we can see, for instance, in Whitelaw’s claim that “the unit of inheritance, i.e., a gene extends beyond the sequence to epigenetic modifications of that sequence” (in: Pennisi 2001: 1064). The molecular process gene concept has, however, a number of potentially troublesome consequences (Moss 2001). First, it substantially increases the number of genes in eukaryotes, due to the great number of polypeptide isoforms generated by alternative splicing. Second, it makes it necessary to include in genes the multimolecular systems associated with transcription and splicing. Thus, the process molecular gene would jump to a higher level in the biological hierarchy. Third, it is hard to individuate genes in accordance with this concept, given the extreme context-dependence of gene expression. According to Moss, the molecular process gene concept falls victim to a ‘conflationary temptation’, which leads to “… an explosion of complexity and contingency that obviates the possibility of a manageable, let alone perspicuous, taxonomy” (Moss 2001: 90). We will return to these potential problems later, when discussing the radical interpretation of our semiotic analysis of the genetic information system (see Ch. 9).
Gene-P and gene-D Another attempt to organize the variety of gene concepts is found in Moss’ (2001, 2003) distinction between gene-P (the gene as a determinant of phenotypes or phenotypic differences, with no requirements regarding specific molecular sequence nor with respect to the biology involved in the production of the phenotype) and gene-D (the gene as a developmental resource which is in itself indeterminate with respect to phenotype).31 Moss forcefully argues that genes can be productively conceived in these two different ways, but nothing good results from previous paper (El-Hani, Queiroz, Emmeche 2006), and is expanded here so as to encompass Fogle’s interpretation of genes as sets of domains (see below). 31 It is important to avoid losing from sight that the distinction between gene-P and gene-D is not identical to the distinction between classical and molecular genes. Molecular entities can be treated as genes-P. For details, see the original works.
66
Genes, Information, and Semiosis
their conflation (Moss 2001: 85), since it would give rise to genetic determinism. Gene-P is the “… expression of a kind of instrumental preformationism” (Moss 2001: 87), showing its usefulness due to the epistemic value of its predictive power and its role in some explanatory games of genetics and molecular biology.32 Gene-D is conceived, in a more realist tone, as a developmental resource defined by a specific molecular sequence and functional template capacity. Gene-D plays an entirely different explanatory game, in comparison to that of geneP. Gene-P and gene-D are, in short, distinct concepts with different conditions of satisfaction for what it means to be a gene. Keller (2000), in turn, argues that we should think of the gene as two different kinds of entities: one would be a structural entity, maintained by the cell’s machinery, so as to be transmitted faithfully from generation to generation; the other would be a functional entity, which emerges only out of the dynamical interaction between a huge number of elements, being the structural gene only one of these. She stresses, thus, that the function of a structural gene depends not only on its sequence, but also on its genetic context, including both the chromosome structure in which it is embedded (which is subject to developmental regulation and can have significant effects on gene expression), and the developmentally-specific nuclear and cytoplasmic context. It is not clear, however, how this proposal can make us advance in view of the difficulties faced by the interpretation of genes as structural units, in particular (see above). After all, Keller does not tell us how to solve the segmentation problem. Second, her treatment of genes as functional entities faces the same difficulties pointed out by Moss in connection to the molecular process gene concept. Symptomatically, she comes close to a process view of the gene when she claims, paraphrasing a rather biosemiotic question posed by Pattee (‘How does a molecule become a message?’ See Pattee 1969), that we should now ask ‘How does a sequence become a gene?’ 32
Moss does not attack the much criticized construct of the ‘gene for’ one or another phenotypic trait, recognizing its value for some theoretical and empirical tasks. This is an interesting new approach to the criticism of the expression ‘gene for’, which is widely seen as the source of risky misunderstandings in the language of genetics and molecular biology (see, e. g., Judson 2001). If we take into account Moss’ distinction between gene-P and gene-D, the danger does not come from this expression in itself but from combining with it a conflation between the two gene concepts.
Chapter 2. The problem of the gene
67
Genes as sets of domains in DNA It is also worth examining a proposal which we regard as both very interesting and somewhat neglected33, namely, Fogle’s proposal of treating genes as sets of domains in DNA. Fogle (1990: 351) argues that we should abandon the classical unit concept and recognize that a gene is constructed from an assemblage of embedded, tandem, and overlapping domains in DNA, if we want to take in due account the complexity and diversity of gene architecture and dynamics. The basic idea is to employ set structure in order to leave aside the need to find a single unit for a region of genetic information. He argues that a gene looks like “… a collection of component entities that together define its structure and influence the phenotype” (Fogle 1990: 367). A gene is better described as a set of ‘domains’, i.e., sequences of nucleotides which can be distinguished from each other on the basis of their structural properties and/or activities: exons, introns, promoters, enhancers, operators, leader and trailer sequences, etc. Domains can be combined in a variety of ways to form sets, or, as Fogle (1990, 2000) calls them, Domain Sets for Active Transcription (DSAT). He finds a similarity between his proposal and procedures used by molecular geneticists, who, when speaking about enhancers, promoters, exons, etc., would have in mind common properties of structure or activity among genes, i.e., they would be “… dissecting genes into domains” (Fogle 1990: 368). It is not necessary, however, to specify all domains influencing expression for nominative or heuristic purposes, but just “sufficient itemization to prescribe a set that has communal agreement” (Fogle 1990: 368, emphasis added). Despite Fogle’s negative appraisal of instrumentalist views about genes, we believe that his proposal suggests an instrumental approach to genes: DSATs seem to be constructs established by agreement in a community of researchers, and it would be quite hard to argue for a hypothesis of correspondence between these constructs and real entities, as we would face again the host of problems discussed above. To be sure, there is a clear realist side to Fogle’s proposal: he treats domains in a realist way, assuming hypotheses of correspondence 33
Among 9 articles which quoted Fogle (1990), according to a survey we performed in Web of Science in January 2005 (http://www.isinet.com/), only Waters (2000) mentions the idea that genes might be treated as sets of domains, but to ultimately reject it.
68
Genes, Information, and Semiosis
between concepts like exons or enhancers and actual structures in DNA. But, when building sets of domains, we are, to use Falk’s (1986: 164) words, superimposing “structure and organization upon the aggregates”. As epistemic objects (Rheinberger 2000), DSATs would be built by communities of researchers in order to manipulate empirical data and organize theoretical knowledge in their fields.34 Fogle’s approach can arguably offer an invaluable contribution to current research in genomics and molecular genetics. Consider, for instance, that, when the draft of the human genome sequence was published in 2001, one of its puzzling features was that our species seemed to have fewer genes than formerly expected. This can be explained by mechanisms which allowed for the expansion of the proteome of metazoans without a corresponding expansion of the number of coding regions in DNA, such as alternative splicing (Maniatis, Tasic 2002). In view of these mechanisms, to count genes in an organism can be very misleading. But to count, instead, the number of sets of domains or DSATs can arguably give a more accurate picture. As Fogle (1990: 368) comments, “[…] the number of DSATs in the genome far exceeds the number of genes as defined by units and more accurately quantifies the number of primary polypeptide products”. Moreover, he argues that “[…] knowing the number of ‘genes’ is a less useful parameter than knowing some measure of the density, type, and number of domains present in a given organism” (Fogle 1990: 369). There are certainly several important tasks to be addressed before contributions can stem from this approach. The expansion of the zoo of instrumentally formulated genetic entities in the last three decades has resulted in a rather loose and sometimes confusing usage of terminology in molecular genetics (Falk 1986). For instance, what is the difference between ‘regulatory element’, ‘cis-acting element’, ‘cisacting sequence’, ‘regulatory sequence’, and ‘5’ regulator’? It is quite clear that these terms substantially overlap in meaning, and most of 34 Using Moss’ distinction, this applies particularly to genes-D. If we treat genes-D as epistemic objects, they will also be instrumental constructs, just as genes-P, as tools for prediction, are. Nevertheless, domains in DNA, which can be combined in a variety of ways to form genes, bring with them a hypothesis of existence in reality, and, therefore, are seen from a realist perspective. If we interpret Fogle’s ideas as we do here, we will end up, thus, with a curious juxtaposition of instrumental and realistic entities in our understanding of genes-D.
Chapter 2. The problem of the gene
69
them should be either eliminated or defined in more precise terms. As another example, we can mention a picture in Chapter 1 of Genomes, a textbook by Brown (2002)35, which shows the location of ‘genes’, ‘gene segments’, ‘pseudogenes’, ‘genome-wide repeats’, and ‘microsatellites’ in a segment of the human β T-cell receptor locus on chromosome 7 containing only 50-Kb. To pose just one question, what is precisely the difference between ‘genes’ and ‘gene segments’? Obviously, this problem cannot be solved merely by interpreting genes as sets of domains, but this interpretation can help by demanding that domains be clearly specified by structure and/or activity. Therefore, a first task is to build a formal system to designate and describe domains in DNA. This can be done through efforts similar to the Gene Ontology Consortium (Ashburner et al. 2000)36, and it would be part of an endeavor that, as Collins and colleagues (2003) predict, seems to be required for the future of research in genetics, molecular biology, and genomics, namely, the development of a comprehensive and comprehensible catalogue of all of the components encoded in the human genome, as well as in other genomes. Nevertheless, it is not the case, we believe, of building a whole new vocabulary, as Brosius and Gould (1992) proposed, based on the perception that genomic nomenclature has not kept pace with the advances in the understanding of genomic structure, function, and evolution. With the purpose of contributing to the integrated study of evolution and molecular biology, they proposed a general terminology based on the introduction of the term ‘nuon’ to name all identifiable structures represented by a nucleic acid sequence (DNA or RNA), for 35 This book can be accessed in United States’ National Center for Biotechnology Information (NCBI), at http://www.ncbi.nlm.nih.gov/books/bv.fcgi?call=bv.View..ShowTOC&rid= genomes.TOC&depth=2. 36 The Gene Ontology Consortium was formed to address the problem that current systems of nomenclature for genes and their products are quite divergent, notwithstanding perceived similarities, holding back the interoperability of genomic databases (Ashburner et al. 2000: 25). The goal of the Consortium was described as follows: “[...] to produce a structured, precisely defined, common, controlled vocabulary for describing the roles of genes and gene products in any organism” (Ashburner et al. 2000: 26). From the standpoint of a view of genes as sets of domains, the divergence of nomenclature for domains in DNA demands, similarly, the construction of a structured and precise system for describing the roles of domains in DNA, and, accordingly, name them.
70
Genes, Information, and Semiosis
instance, genes, intergenic regions, exons, introns, promoters, terminators, pseudogenes, transposons, etc. It is clear that the term ‘nuon’ plays in his scheme the same role played by the term ‘domain’ in Fogle’s. Nevertheless, they go far beyond in their proposal of a new terminology. They suggest that, when a ‘nuon’ is clearly defined (structurally, functionally, or by its origin), it can receive a more specific name by the use of a prefix. They do not propose to rename all existing elements, for example, to call a promoter a ‘promonuon’, but they indeed propose the inclusion of new terms in a number of cases. To address the process of gene amplification, for instance, they propose that any duplicated or amplified gene or nuon be called a ‘potogene’ or ‘potonuon’ (the prefix ‘pot’ standing for ‘potential’). If a potonuon never acquires a function in the course of evolution, eventually becoming only ‘genomic noise’, it can be termed a ‘nonaptive’ nuon or ‘naptonuon’. In contrast, if a nuon is coopted for a function, i.e., if it turns out to be an exaptation (Gould, Vrba 1982), it can be called a ‘xaptonuon’ or ‘xaptogene’. Not surprisingly, their perplexing proposal was entirely ignored, as they themselves anticipated. The second development from Fogle’s ideas would be to establish formal procedures for the combinations of domains in DSATs or genes, taking in due account, as far as possible, the practices currently used by the communities of geneticists and molecular biologists. After all, they would ultimately have to make use of the ‘libraries’ of domains and formal rules of combination resulting from such an effort.
The systemic gene concept Pardini and Guimarães (1992) also argue against the unit concept and propose, instead, a ‘systemic concept of the gene’, according to which the gene is a combination of (one or more) nucleic acid (DNA or RNA) sequences, defined by the system (the whole cell, interacting with the environment, or the environment alone, in subcellular or pre-cellular systems), that corresponds to a product (RNA or polypeptide). (Pardini, Guimarães (1992: 717, see also 713)
This definition treats the genome as part of the cellular system, which “builds, defines and uses the genome as part of its memory mechanisms, as an interactive database” (Guimarães, Moreira 2000: 249).
Chapter 2. The problem of the gene
71
Two kinds of correspondence between the inherited nucleic acid segments and their products, and, consequently, two types of genes can be distinguished, according to Pardini and Guimarães: a univocal type, predominant in bacteria, in which a one-to-one correspondence between nucleic acid sequences and gene products holds, and only gene regulation depends on the system in which those sequences are embedded; and a multivocal type, more common in eukaryotes, in which the defining role of the system increases, affecting both the structure and function of the gene.37 The systemic concept of the gene shows some similarities to Fogle’s conception of genes as sets of domains in DNA, even though Fogle’s proposal seems to be restricted to DNA genomes (Guimarães, pers. comm.). Pardini and Guimarães (1992: 716) stress the dynamics of the relationship between encoded information and the product of its decoding, which is quite complex, varying with the spatial and temporal conditions of occurrence. Guimarães and Moreira (2000) argue that the meaning of a DNA segment is relative, depending on the expression system in which it is embedded. Consequently, its meaning can be plural: the multivocal nature of genes, particularly in eukaryotes, stems from the context-dependence of gene expression. Even though it seems tempting to conclude that this approach is similar to Neumann-Held’s process molecular gene concept, it is important to emphasize that this would be a mistake (Guimarães, pers. comm.). The systemic gene concept indeed alludes to the process which specifies or demarcates the gene, as an adequate combination of (one or more) genomic sequences corresponding to a product, but it takes this demarcation as the result of a process which is not included itself in the gene. Pardini and Guimarães (1992) discuss the consequences of their approach to the detection of genes in nucleic acid sequences, arguing that a functional test (in vivo or in vitro) is needed to say whether a given nucleic acid sequence is a gene or not. This idea implies that it is not possible to deduce genes simply from nucleotide sequences, as it is currently widely acknowledged, and also that the status of sequences as genes will be relative to the systems in which they successfully work. 37 Guimarães and Moreira (2000) discuss different hypotheses for the evolution of multivocal genes.
72
Genes, Information, and Semiosis
They also highlight the didactic value of their approach by critically appraising Lewin’s (1990) idea that the problems with the unit concept might be overcome by moving from a ‘one gene-one polypeptide’ to a ‘one polypeptide-one gene’ correspondence rule.38 For them, this is only “… a technical solution, adequate to the observer’s needs” (Pardini, Guimarães 1992: 718, emphasis added), committed as it is to the classical Mendelian approach of inferring inheritance units from the phenotype. They assume a more realist view about the gene, which is particularly explicit in their statement that “if a concept intends to be used in full for didactic, heuristic and research purposes, it should try to portray or reflect reality faithfully” (ibid., p 719). Epistemologically speaking, this is certainly a polemical statement, but it suggests, anyway, that, despite the return of an instrumentalist attitude towards the gene (Falk 1986), we are likely to find among geneticists and molecular biologists a lasting commitment to realist intuitions about the theoretical concepts they work with. In a reassessment of the systemic gene concept, Guimarães and Moreira (2000) introduce, as the main novelty, the understanding of the system which defines the genes as a ‘hypertext’.39 In an analogical model, they describe the system as an encyclopedia, in which genes are headings of the entries and the explanatory bodies are hypertexts, which emerge from the interactions between gene products, through which the metabolic web is organized. As suggested immediately by this analogy, they show sympathy towards biosemiotics, arguing that discussions about the gene concept should ‘obviously’ flows to a biosemiotic approach. We do not think, however, that the particular analogical model they propose is adequate to address the nature of genomic systems and their relationship to cellular and supracellular systems. 38
This proposal is also discussed by Fogle (1990). They also differentiate between two concepts of gene: the classical molecular concept, in which humans are interested as technological agents, and the systemic gene concept, as a ‘natural’ system, which is demanded for the description of how the cellular system functions (Guimarães, Moreira 2000). We are not convinced that it is worth maintaining this distinction, since the technological manipulation of genetic systems itself cannot and should not ignore the complexities of genes and their function in cellular systems, as the classical molecular concept prompts us to do. 39
Chapter 2. The problem of the gene
73
11. But… what about ‘information’? Fogle (1990: 368, emphasis added) observes that “a domain model accepts that information resides within units smaller or larger than a conventionally defined gene and that sets of domains are needed to establish correspondence with Mendelian entities”. It is clear that Fogle is still committed to an interpretation of ‘information’ as something carried by stretches of DNA. The same is true of other accounts, such as, for instance, Pardini and Guimarães’ (1992: 714), who argue that “… biological information is very densely packed in some segments of the inherited genomes” (emphasis added), or, else, that “the genome contains information, as its main depository, but each function or expression has to be interpreted from it” (Pardini, Guimarães 1992: 716).40 This interpretation of ‘information’ is challenged in this book, and, as we will see in the next chapters, in our attempt to do so, we will be even taken as far as Neumann-Held’s claim that genes should be understood as processes. We will end up with a pair of alternative interpretations: in one of them, genes will be treated as sets of domains in DNA which are interpreted, built, demarcated, and used by the cellular system as an interactive set of signs, much in line with Fogle’s and Pardini, Guimarães’ proposals, but these sets will not be treated as carriers of genetic information (which will be itself regarded as a process); in the other, genes will be treated as the entire molecular process underlying the capacity to express a particular product, in accordance with the process molecular gene concept. In the next chapter, we will turn to discussion about information talk in biology and a possible role for biosemiotics in building a theory of biological information, which can allow us to clarify, if not go beyond, metaphorical usage of the informational vocabulary currently used in this science.
40
But notice, in the case of this passage, an opening for a semiotic treatment of genetic information, as something which is not simply deposited in the genome, but should be interpreted from it.
Chapter 3 Biosemiotics and information talk in biology
‘Information’ is an important but problematic concept in biology (see Oyama 2000; Stuart 1985; Hoffmeyer, Emmeche 1991; Sarkar 1996; Griffiths 2001; Jablonka 2002). This concept has been recently a topic of substantial discussion (see, e. g., Guimarães 1998; Maynard Smith 2000a; Godfrey-Smith 2000; Sarkar 2000; Sterelny 2000; Wynnie 2000; Jablonka 2002; Adami 2004). The evolution of new kinds of information and information interpretation systems in living beings has also received a great deal of attention recently (See, e. g., Jablonka 1994; Jablonka, Szathmáry 1995; Maynard Smith, Szathmáry 1995, 1999; Jablonka, Lamb, Avital 1998; Jablonka, Lamb 2005). It is even the case that the evolution of different ways of storing, transmitting, and interpreting ‘information’ can be treated as a major theme in the history of life (Maynard Smith, Szathmáry 1995, 1999; Jablonka 2002). Nevertheless, it is not clear at all what is meant by ‘information’ in the biological sciences, particularly in fields which have been swamped, during the 1950s and 1960s, by terms borrowed from information theory, such as genetics and cell and molecular biology.41 ‘Information talk’ still pervades these fields, including widely used terms such as ‘genetic code’, ‘messenger RNA’, ‘transcription’, ‘translation’, ‘transduction’, ‘genetic information’, ‘chemical signals’, ‘cell signaling’, etc. But several conceptual problems resulted from the introduction of ‘information’ and its plethora of associated notions, and the tradition of biology didn’t have the theoretical and methodological resources to cope with them. Instead of deepening the discussion about the problems involved in information talk, the trend in genetics and molecular biology was one of treating ‘information’ as merely sequence information in DNA or proteins (Emmeche, Hoffmeyer 1991; Hoffmeyer, Emmeche 1991; Sarkar 1998a). Nevertheless, this interpretation can and should be challenged, as we will do throughout this book. Some years ago, a target paper about the concept of information in biology was published in Philosophy of Science, focused mainly on genetic information. In this paper, Maynard Smith (2000a) claimed that the notion of biological information is closely related to the idea 41 For a historical account about how the genome became an information system and DNA a language, see Kay (1995, 2000).
78
Genes, Information, and Semiosis
that both the signal carrying the information and the response to it are products of natural selection. However, interesting as they are, Maynard Smith’s arguments do not go a long way toward clarifying the basic concept at stake: what, after all, is ‘genetic information’? As Griffiths (2001) wrote recently, ‘genetic information’ is still a metaphor in search of a theory. We think the same can be said, in general terms, of the current use of the term ‘information’ and its derivatives in biology. After all, a number of researchers consider all information-talk as inadequate, taking a skeptical view toward the very use of the term ‘information’ and its derivatives in biology as a natural science (e. g., Stuart 1985; Sarkar 1996). Among the reasons for this skepticism is the fact that the use of that term in biology is not as precise as in the mathematical theory of communication. Furthermore, although the standard account of genetic information refers to an alleged semantic property of genes, it is not clear if and how any genuine semantics is involved. One possibility for building a theory of information in biology is to rely on the mathematical theory of communication. This theory is a branch of mathematics that arose out of communication theory. As Shannon and Weaver defined it, “[t]he fundamental problem of communication is that of reproducing at one point either exactly or approximately a message selected at another point” (1949: 31). According to Adams: at the foundation of information theory is the development of methods to measure the amount of information generated by an event or events, and mathematical treatments of the transmission characteristics of communication channels. (Adams 2003: 472)
It relies on the theory of probability to model information sources, flow, and communication channels. The amount of information is measured in terms of the unexpectedness of the sequence of signals, i.e., the mean amount of information per signal can be written H=∑ pi log (1/pi), where pi is the probability of the ith form of a signal. This theory allows one to define the amount of information as the measure of the probability of selection of a particular message among the set of all possible messages. The probabilistic measure of information provided by this theory is non-semantic, indifferent to meaning (Shannon, Weaver 1949; Cover and Thomas 1999; Jablonka 2002). Despite the fact that this meaning-free concept of information can be
Chapter 3. Biosemiotics and information talk in biology
79
useful in biological research for several purposes (Adami 2004), polemics continues about a non-semantic understanding of information in this science. In particular, it is controversial whether such an understanding is sufficient for a theory of biological information. Jablonka (2002), for instance, argues that this concept is not sufficient by pointing out that, for instance, a DNA sequence encoding a functional and a same-length sequence coding for a completely non-functional enzyme (which can even have only a single different nucleotide) would contain, according to the above-mentioned measure, the same amount of information. It is obvious, however, that these two messages do not mean the same to the cell. This indicates the necessity of a definition of information in biology which includes a semantic and a pragmatic dimension. Or, to put it differently, in living systems the focus of a theory of information should be on the meaning of ‘messages’ and the context in which they are interpreted. Nevertheless, while in the case of the gene, a number of definitions have been coined and discussed, semantic and pragmatic concepts of information have been rarely defined in biology (Jablonka 2002). Moreover, several authors, particularly Susan Oyama (2000), argued that the usual way of applying the concept of information to biological systems raises a number of important problems. Oyama argues that genetic determinism is inherent to the way we represent genes and their function in biological systems as if they carried information about how an organism will develop, as containers of developmental information. If this conception continues to prevail, she argues, genes will continue to be regarded as deterministic and centrally directing causes, no matter how much evidence against this idea exists. As we mentioned above, Stuart (1985) and Sarkar (1996) proposed that information talk should be eliminated from biology, since ‘information’ is a foreign metaphor in this science and its use may lead to erroneous views of explanation in fields such as molecular biology. Griffiths and Neumann-Held, in turn, argue that the metaphorical ideas of DNA containing the ‘information for’, ‘coding for’, or ‘programming for’ phenotypic traits cannot be the final stopping points in an explanation of development. […]. Indeed, the idea that developmental biology could be replaced by going straight to the genes and reading the instructions for development has been rightly mocked as ‘neo-preformationism’. (Griffiths, Neumann-Held 1999: 656)
80
Genes, Information, and Semiosis
These worries are justified, and, accordingly, to build a theory of information in biology, it is fundamental to avoid deterministic and gene-centric ideas which have marked the informational conception of the gene throughout its history. It is also necessary to elaborate a precise account of what is ‘information’ in biology. Genetics and molecular biology are committed to a rather loose talk about information. Griffiths and Neumann-Held (1999: 657), for instance, observe that “the conventional defense of […] informal “information talk” in genetics is that it is meant ‘more or less in the spirit of information theory’.”42 They argue, then, that “the possibility of translating the information metaphor into substantive theory is an illusion”, since “repeated attempts to bridge the gap between information talk and information theory in molecular biology have been unsuccessful”. Indeed, any attempt to defend information-talk in terms like those employed by Maynard Smith is untenable. But it may be too hasty to declare that it is nothing but an illusion to try to translate ‘information’ in biology and related concepts into a substantive theory on the basis of attempts to tackle the issue by taking information theory as a theoretical ground. After all, the non-semantic understanding of information in this theory can arguably be the reason why previous attempts to build a theory of bio-information failed. One can argue — as we do in this book — that the task of going beyond information metaphors in biology and building a theory of information in this science demands a semantic and pragmatic account, and, accordingly, the right tools to accomplish it. From a biosemiotic perspective, we think that a semantic and pragmatic notion of information and its derivatives cannot be dispensed with, since they grasp some fundamental features of biological systems and processes that might be otherwise neglected. In particular, the concepts of ‘code’, ‘information’, ‘signals’, ‘message’, ‘signaling’, ‘transduction’ and so on seem to be necessary for an understanding of the organization of relations in living beings which make it clear that what happens in such beings is much more than simple chemistry (for details, see Emmeche, Hoffmeyer 1991; Emmeche 1991). For instance, understanding control and regulation in cellular systems without understanding cell signaling is simply not 42
They are quoting a statement by John Maynard Smith, in the Fifth International Congress of Systematics and Evolutionary Biology, held in Budapest, 17– 24 August 1996.
Chapter 3. Biosemiotics and information talk in biology
81
possible. Moreover, it is the spatial and temporal coordination of processes that makes a cell a living system, and it crucially depends on control and regulation. This is the reason why ‘systems biologists’ in genomics and molecular biology came to realize — rather late, we should say — that to understand the dynamics and structure of genes, we should locate them in complex informational networks and pathways, which are ultimately responsible for controlling the timing of gene expression in cells. As a consequence, biology is even described from this standpoint as an ‘information science’ (Ideker et al. 2001). This shows both that information talk is still on the crest of the wave in molecular biological research, but also that the problem of building a theory to say what, after all, is information in biology remains. Biology has no theory at present to back up these rhetorical, sometimes even far-fetched claims. But we do not think it is the case of casting away the claims; rather, we should build a theory to support them! These statements show that molecular and cell biologists, geneticists, etc. harbor an intuition that the concept of information plays a fundamental role in their sciences, and this role is closely linked to a non-reductionistic belief that by taking into due account informational networks we will come to understand how cells are more than merely sacs of molecules. More examples can be quoted. Bray, in a symposium about reductionism in 1997, argued that, since “about 50% of the genome of a multicellular organism may code for proteins involved in cell signaling, [...] organisms can be viewed as complex informationprocessing systems, where molecular analysis alone may not be sufficient” (cited by Williams 1997: 476–477). Similarly, Nurse argued that “there’s a need to realize that information may be transmitted in ways that may be lost by studying molecules alone”, and, furthermore, that “it may not be possible or even necessary to explain all cellular phenomena in terms of precise molecular interactions” (cited by Williams 1997: 476–477).43 These statements illustrates the current zeitgeist in molecular biology, in which it is increasingly recognized that more than just chemistry is taking place in living beings because these systems process ‘information’ in quite complex ways. In other words, biological 43
By the way, Nurse’s claim is in opposition to the current large-scale reductionistic projects of cataloguing all molecules and their interactions within a living cell (Barabási, Oltvai 2004: 101. See Chapter 2).
82
Genes, Information, and Semiosis
meaningfulness is emerging all the time in such systems. It is not surprising that biologists feel the need to talk about ‘information’ as they delve more and more in the molecular micro-structure of living systems. It is just the case that they need a way of conveying the idea that, even though all cellular processes are physicochemical processes, more than just physics and chemistry is going on there (and is nothing like any vital principle, energy, or élan vital). From this perspective, it is quite difficult to see what would be the real advantage of stripping off biology of information talk, instead of making it more precise and exploring its consequences further. The task of building a theory of biological information becomes more and more important as our knowledge about the structural and functional complexity of living beings increases. The concept of information and related notions should not only be taken seriously in biology, but also clarified by employing appropriate conceptual tools. Biosemiotics offers us a way of facing this challenge, by more precisely explaining in what sense informational notions can show how something physical can give rise to processes which cannot be explained as being purely physical. In this manner, biosemiotics can contribute to the construction of a precise and coherent formulation of the notion of information in biology. A semiotic treatment of information talk can significantly add, for instance, to an understanding of the role of genes in biological systems which avoids the reference to notions much criticized such as ‘blueprints’ and ‘programs’, while preserving the concept of ‘information’, albeit radically reinterpreted. It also lends support to the now widely accepted idea that there is more to information in living systems than just genes (see, e. g., Jablonka 2002). In this book, we will develop a Peircean semiotic account of genes as signs and, simultaneously, propose a new understanding of information in biological systems. Both steps are fundamental, in our view, to the construction of a theory of biological information. It is important to stress, however, that there is no consensus about the notion of information within the biosemiotics community. Therefore, our readers should bear in mind that the purpose of this book is to contribute to an account of information from the perspective of Peircean biosemiotics, and other perspectives on biosemiotics might elaborate different accounts.
Chapter 3. Biosemiotics and information talk in biology
83
Finally, we should emphasize that we concentrate our efforts in this book on genetic information simply for methodological reasons. We certainly recognize that there are several other information systems in living beings. Jablonka (2002), for instance, describes five different information systems in living organisms, each involving particular modes of inheritance or ‘information re-production’: genetic, epigenetic cellular44, epigenetic organismic45, behavioral, and (human) symbolic46 information systems.47 Accordingly, inherited information is not limited to genetic information. The reason why we take the genetic information system as a starting-point for a semiotic treatment of biological information is related, first, to its important role in biological thinking, and, secondly, to the fact that it offers some clear-cut case studies for the building of a toolkit for semiotic 44
The epigenetic cellular information system includes, for instance, self-sustaining regulatory loops, through which the state of activity of a metabolic network is maintained via positive feedback; templating of molecular structures, in which 3-D cellular structures act as templates for the production of new daughter structures, as in the case of prions and membranes; chromatin marking, based on patterns of DNA-bound molecules, such as proteins, RNAs, and small chemical groups (e. g., methyl groups); RNA-mediated gene silencing, in which small RNA molecules (siRNAs) act so as to suppress the expression of a given gene. For more details, see Jablonka, Lamb, Avital (1998); Jablonka (2001, 2002). 45 The epigenetic organismic information system concerns ‘developmental legacies’, molecular products and consequences of the parental developmental history that can sometimes be re-produced and regenerated in the progeny, as in the case of the inheritance of gender behavior and sex ratio in Mongolian gerbils (Clark, Karpiuk, Galef Jr. 1993) or the inheritance of food and host preferences in lineages of insects (Thompson, Pellmyr 1991). See Griesemer (2000), Jablonka (2001, 2002). 46 The use of the concept of ‘symbol’ by Jablonka (2002) indicates another set of issues in which a semiotic framework can be of great help. It is even possible that symbols, as interpreted from a Peircean perspective, appear in non-human animals (Ribeiro et al. 2007, Queiroz, Ribeiro 2002, Queiroz 2003). It is beyond the scope of this book, however, to consider this point and we should leave it to future works. 47 For the details of this analysis, see the original paper. See also Jablonka, Lamb, Avital (1998); Jablonka (2001). Yet another cellular information system deserves investigation, based on what has been called the ‘sugar code’, established by monosaccharides and decoded by the interplay of sugar receptors, including lectins, enzymes, and antibodies, so as to trigger signaling and elicit biological responses. For a review on the sugar code, see Gabius (2000). Two other information systems also fall under the scope of semiotic analyses, namely, the nervous system and the immune system.
84
Genes, Information, and Semiosis
analyses of information systems in living beings. We do not intend, however, to take the genetic information system as a prototype for other biological information systems. As Jablonka (2002: 579) argues, despite its importance, the genetic system is highly specific and unusual, and, therefore, should not be taken as a prototype for thinking about information in biology in general terms. We consider the semiotic analysis of the genetic information system presented in this book just a first step in a research program aiming at a general semiotic analysis of information systems in living beings. Our intention is to proceed in subsequent works with semiotic analyses of other biological information systems, using the theoretical framework built during the analysis of the genetic information system, but adapted to each specific case under analysis. We hope this will show both the similarities and differences between the diverse information systems with which biology should deal to be truly — and not only rhetorically — an ‘information science’, capable of explaining how physical and chemical entities can be informative to the organism, not only to scientists.48
48
Cf. Bruni’s (2003) argument that in the proposal that biology is an information science (which is gaining currency in molecular biology and genomics) we find a systematic confusion between information handled by the organism and information used by researchers, as observers of living systems.
Chapter 4 Information and semiosis in Peirce’s science of signs
1. Introduction When Peircean semiotics is used as a theoretical framework for case studies about specific meaning processes in biology, one should remember that the notion of Sign in Peirce is not the same as a simple ‘unit’ of information or communication as these terms are often used in several fields of research. It is a notion related to formal attempts to describe mind processes in general (Skagestad 2004). It is our primary aim here to apply some central notions of Peirce’s semiotics to understand the nature of genetic information. Nevertheless, such an application necessarily involves interpretation and, thus, decisions about how to see, for instance, the relationship between what molecular biologists and geneticists call forms of information processing (i.e., production and interpretation of Signs) in a complex living system such as the cell and forms of causality in that system. The analysis of the genetic information system given below is obviously not the only way to apply Peircean semiotics to this particular case; and some might object to the particular way we addressed the problem. Anyway, we think we have been faithful to both the basic insights and concepts of semiotics, and the findings of molecular biology, and the few changes we made in specific semiotic conceptions (as we shall explicate below) are necessitated by the growth of scientific knowledge about the system analyzed.
2. Some basic ideas in Peirce’s semiotics Peirce’s conception of Semiotics as the ‘formal science of signs’ has had a deep impact on philosophy, psychology, theoretical biology, and cognitive science (see Stjernfelt 2007; Short 2007; Gudwin, Queiroz 2007; Pietarinen 2005; Queiroz, Merrell 2005; Petrilli, Ponzio 2005; Freadman 2004; Shin 2002; Hookway 2002; Violi 1999; Houser 1997; Deacon 1997; Brunning, Forster 1997; Hoffmeyer 1996; Tiercelin 1995; Colapietro 1989; Freeman 1983, Jakobson 1969, 1973). Peircean semiotics is based on a theory of categories, including a list of categories (Firstness, Secondness, Thirdness) which can be logically
88
Genes, Information, and Semiosis
described as an exhaustive system of hierarchically organized classes of relations (monadic, dyadic, triadic) (Houser 1997; Brunning 1997). This system is the formal foundation of his ‘architectonic philosophy’ (Parker 1998) and of his model of semiosis (Sign action) (Murphey 1993: 303–306). Briefly, the categories can be defined as follows: Firstness: what is such as it is, without reference to anything else; Secondness: what is such as it is, in relation with something else; Thirdness: what is such as it is or becomes, insofar as it is capable of bringing a second entity into relation with a first. Firstness is the category of mere potentiality, freedom, immediacy, undifferentiated quality, randomness, independence, novelty, creativity, and originality. Secondness is the category of action and reaction, opposition, polarity, differentiation, existence. Thirdness is the category of mediation, law, habit, semiosis, representation. (for further details on the categories, see Hookway 1985; Murphey 1993; Potter 1997; Merrell 1995). Accordingly, Peirce defined semiosis as an irreducible triadic relation between Sign-Object-Interpretant (S-O-I) (see Savan 1988; Hookway 1992: 121). That is, according to Peirce, any description of semiosis involves a relation constituted by three irreducibly connected terms, which are its minimal constitutive elements (MS 318:81): My definition of a sign is: A Sign is a Cognizable that, on the one hand, is so determined (i.e., specialized, bestimmt) by something other than itself, called its Object, while, on the other hand, it so determines some actual or potential Mind, the determination whereof I term the Interpretant created by the Sign, that that Interpreting Mind is therein determined mediately by the Object. (CP 8.177. Emphases in the original).49
Peirce conceives a ‘Sign’ or ‘Representamen’ as a ‘First’ which stands in such a genuine triadic relation to a ‘Second’, called its ‘Object’, so as to be capable of ‘determining a Third’, called its ‘Interpretant’, to assume the same triadic relation to its Object in which it stands itself 49
Peirce gave a whole family of similar definitions of the Sign, but at this stage of the argument we only intend to give the reader a first idea, and not to be distracted by the question of what Peirce in the quote meant by ‘a Cognizable’, as in other definitions he simply stated that part as “a sign is ‘something’ (or a ‘First’) that determines …” (and so on), not specifying that something’s cognitive or material status. The quoted definition could be restated formally as “a Sign is an R that is so determined by an O to mediately determine an I to stand in a similar relation to O as R itself stand”, as the reader will see below.
Chapter 4. Information and semiosis in Peirce’s science of signs 89
to the same Object (CP 2.274. See also CP 2.303, 2.92, 1.541). The term ‘Sign’ was used by Peirce to designate the irreducible triadic relation between S, O and I as well as to refer to the first term of the triad (sometimes ‘Representamen’). Some commentators proposed, then, that we should distinguish between ‘Sign in a broad sense’ and ‘Sign in strict sense’ (e.g, Johansen 1993: 62). Charles Morris (see Nöth 1995: 80) proposed the use of the expression ‘Sign vehicle’ in the place of ‘Sign in strict sense’. We will systematically use the term ‘Sign’ in this book to refer to the first term of the triad, and ‘semiosis’, to refer to the whole triad. We will not use the expression ‘Sign vehicle’ often but sometimes we will employ it due to its interesting metaphorical connotations in the case of biological systems, in which the first term of a triadic-dependent process is typically a physical entity, such as a molecule or set of molecules. In these cases, we apply the notion of ‘Sign vehicle’ especially to emphasize the ‘material quality’ of a Sign (CP 5.287). Semiosis was regarded by Peirce as irreducible, in the sense that it is not decomposable into any simpler relation:50 ... by ‘semiosis’ I mean [...] an action, or influence, which is, or involves, a cooperation of three subjects, such as a sign, its object, and its interpretant, this tri-relative influence not being in any way resolvable into actions between pairs. (CP 5.484)
One of the most remarkable characteristics of Peirce’s theory of Signs is its dynamical nature (Queiroz, Merrell 2006). According to Merrell (1995: 78), “Peirce’s emphasis rests not on content, essence, or substance, but, more properly, on dynamic relations. Events, not things, are highlighted.” The complex S-O-I is the focal-factor of a dynamical process (Hausman 1993: 72). Peirce was a truly process thinker (see Rescher 1996). It is also important to avoid losing from sight the distinction between the interpreter, which is the system which interprets the Sign, and the Interpretant. The interpreter is described by Peirce as a ‘Quasimind’ (CP 4.536), a description which demands, for its proper interpretation, a clear recognition of Peirce’s broad concept of ‘mind’ (Ransdell 1977; Santaella-Braga 1994). It is far from being the case 50
About the demonstration of the irreducibility of a triadic relation, see Ketner (1986); Brunning (1997); Burch (1991, 1997).
90
Genes, Information, and Semiosis
that only conscious beings can be interpreters in a Peircean framework. Rather, a transcription machinery synthesizing RNA from a string of DNA or a membrane receptor recognizing a given hormone can be regarded as interpreters in such a framework. A basic idea in a Peircean semiotic understanding of living systems is that these systems are interpreters of Signs, i.e., that they are constantly responding to signs in their surroundings. The interpreter does not have to be a conscious being, not even an organism, as it may be some part or subsystem within an organism, or a humanly-designed product.51 To be sure, there is a huge difference between interpretation as accomplished by humans, with their embodied minds as interpretive systems, and, say, interpretation as performed by a cell with membrane receptors. But the basic underlying, triadic logic of semiosis is the same, according to a Peircean framework. Despite all the undeniable differences between diverse interpretation processes in equally diverse interpreters, from a general semiotic perspective it is the triadic relation between Sign, Object, and Interpretant that ultimately allows interpretation to take place. For some ideas about how an evolutionary process of ‘semiotic intensification’ might modify the embodied and intentional aspects of semiosis in different kinds of systems, see Emmeche (2003). A detailed treatment of this issue is beyond the scope of this book.
51
When a part or subsystem of a system is the interpreter, its actions as an interpreter will be typically subordinated, i.e., regulated by the system as a whole (that we will call, in this case, a ‘global’ interpreter). We can call, as Jablonka (2002), the subordinated interpreters ‘interpretative systems’ within a global interpreter. It can happen that a system loses its control over one or more of its included interpreters. In this case, dysfunctional states may result from the interpretation of Signs in that system. It is possible to analyze in these terms, for instance, events in carcinogenesis in which stretches of DNA are transcribed in a place and/or time in which they were not supposed to be transcribed. These would be misinterpretation events. By ‘misinterpretation’, we mean the interpretation of a Sign that does not lead to a successful coping with a system’s circumstances, i.e., that does not contribute to the maintenance of the dynamic stability of a system in a given circumstance.
Chapter 4. Information and semiosis in Peirce’s science of signs 91
3. Kinds of Objects and Interpretants We also need to consider here Peirce’s distinctions regarding the nature of Objects and Interpretants (for a review of these topics, see Savan 1988; Liszka 1990; Short 1996). He distinguishes between the Immediate and the Dynamical Objects of a Sign as follows: We must distinguish between the Immediate Object — i.e., the Object as represented in the sign — and [...] the Dynamical Object, which, from the nature of things, the Sign cannot express, which it can only indicate and leave the interpreter to find out by collateral experience. (CP 8.314. Emphases in the original)
Additionally: [...] we have to distinguish the Immediate Object, which is the Object as the Sign itself represents it, and whose Being is thus dependent upon the Representation of it in the Sign, from the Dynamical Object, which is the Reality which by some means contrives to determine the Sign to its Representation. (CP 4.536)
And we should also take into account his distinction between the following two kinds of interpretants:52 The Immediate Interpretant is the immediate pertinent possible effect in its unanalyzed primitive entirety. […]. The Dynamical Interpretant is the actual effect produced upon a given interpreter on a given occasion in a given stage of his consideration of the Sign. (MS 339d: 546–547, emphasis in the original)
Let us consider, first, Peirce’s distinction between the Immediate and the Dynamic Objects of a Sign. The Immediate Object of a Sign is the Object as it is immediately given to the Sign, the Dynamical Object in its semiotically available form. The Dynamical Object is something in reality which determine the Sign to its representation, and which the Sign can only indicate, something that the interpreter should find out by collateral experience (EP 2.498). In turn, Peirce defines the Dynamical Interpretant as the actual effect of a Sign, while the Immediate Interpretant is its ‘range of interpretability’ — the range of possible effects that a Sign is able to 52 Peirce’s concept of the Final Interpretant will not play a role in our arguments in this book and we leave it to subsequent works.
92
Genes, Information, and Semiosis
produce (see Johansen 1993: 166–167). The Dynamical Interpretant is the instantiation of one of the possible effects established in the Immediate Interpretant. As the effect of the Sign upon the interpreter, the Dynamical Interpretant can be treated as being essentially equal to the significance of the Sign when seen in a dynamic and processoriented perspective.
4. Semiosis, information, and meaning The notions of ‘meaning’, ‘information’, and ‘semiosis’ intersect and overlap in different ways (see Johansen 1993). According to Debrock (1996), Peirce defined ‘information’ at least ordinarily (CP 2.418), metaphysically (CP 2.418), as a connection between form and matter, and logically (W 1.276), as the product of extension and intension of a concept. Here, we systematically refer to information as the communication of a form from O to I through S. In these terms, it amounts to the communication of a habit embodied in the Object to the Interpretant, so as to constrain (in general) the Interpretant as a Sign or (in biological systems) the interpreter’s behavior. We should also stress that by ‘communication’ here we mean more than mere transmission of a form. To put it in more detailed terms, the production of an effect of the Sign on the interpreter results from the communication of the form of the Object (as a regularity), by Sign mediation, to the Interpretant. The Interpretant then becomes itself a Sign which refers to the Object in the same manner in which the original Sign refers to it (i.e., there is an invariance in the reconstruction of the form of the Object by the interpreter). According to this approach, ‘information’ can be strongly associated with the concepts of ‘meaning’ and ‘semiosis’. Peirce spoke of Signs as ‘conveyers’, as a ‘medium’ (MS 793), as ‘embodying meaning’. In short, the function of the Sign is to convey the form (EP 2:391): [...] a Sign may be defined as a Medium for the communication of a Form. [...]. As a medium, the Sign is essentially in a triadic relation, to its Object which determines it, and to its Interpretant which it determines. [...]. That which is communicated from the Object through the Sign to the Interpretant is a Form; that is to say, it is nothing like an existent, but is a power, is the fact that something would happen under certain conditions. (MS 793: 1–3. cf. EP2: 544, n.22)
Chapter 4. Information and semiosis in Peirce’s science of signs 93
But… what is a Form? There is a movement in Peirce’s writings from ‘form as firstness’ to ‘form as thirdness’. Form is defined as having the ‘being of predicate’ (EP 2.544) and it is also pragmatically formulated as a ‘conditional proposition’ stating that certain things would happen under specific circumstances (EP 2.388). It is nothing like a ‘thing’ (Tienne 2003), but something that is embodied in the object (EP 2.544, n. 22) as a habit, a ‘rule of action’ (CP 5.397), a ‘disposition’ (CP 2.170), a ‘real potential’ (EP 2.388) or, simply, a ‘permanence of some relation’ (CP 1.415). We can say that Peirce follows a via media in which ‘form’ has both the characters of firstness and thirdness. This is in accordance with Bergman’s (2000: 236) proposal of communicated form as a First of a Third. He based his proposal on the modalities associated with Firstness (possibility), Secondness (existence), Thirdness (habit, law), and on the principle of the interdependence of categories (see Potter 1997:14). It is particularly important to notice that the form commmunicated from the Object to the Interpretant through the Sign is not a thing or a particular shape of a thing, or something alike, but a regularity, a habit which allows a given system to interpret that form as indicative of a particular class of entities, processes, phenomena, and, thus, to answer to it in a similarly regular, lawful way. Otherwise, the system would not be really capable of interpreting the Object by means of its effect on it (Interpretant), mediated by a Sign. It is possible to understand the idea of ‘interpretation’ as basically meaning to subsume a given particular event under a general class of events, and, by thus subsuming it, to answer to it in a regular way, learnt by systems through evolution or development (i.e., by evolutionary or somatic learning). Peirce defines a Sign, in the passage quoted above, both as ‘a Medium for the communication of a Form’ and as something which is in ‘a triadic relation, to its Object which determines it, and to its Interpretant which it determines.’ If we consider both definitions of a Sign, we can say, then, that semiosis is a triadic process of communication of a form from the Object to the Interpretant by the Sign mediation (Figure 2). Therefore, in this framework, we can say that semiosis is information, if we define this latter concept as above.53
53
Notice that our argument is grounded on a passage found in a late work by Peirce, in which the idea of a Sign as a medium for the communication of a form is prominent (EP2.544, n. 22; also EP2:329, EP2:389–391).
94
Genes, Information, and Semiosis
sign
form object
interpretant
Figure 2. Semiosis as a relation between three irreducibly connected terms (signobject-interpretant, S-O-I). This triadic relationship communicates a form from the object to the interpretant through the sign (symbolized by the horizontal arrow). The other two arrows indicate that the form is communicated from the object to the interpretant through a determination of the sign by the object, and a determination of the interpretant by the sign.
Peirce (see Fitzgerald 1966: 84; Bergman 2000) defined meaning as connected to the triadic relation as a whole (EP 2:429), as well as to different correlates of a triad — e. g., Object (MS 11, EP 2:274), Interpretant (EP 2:496, EP 2:499; CP 4:536). Here, we focus on Peirce’s explanation of meaning as something communicated in semiosis (NEM 4:309), and, therefore, we argue that meaning can be plausibly treated as connected to the interpretant, which embodies a reconstructed form of the Object. We will also consider here Peirce’s (CP 8.177) idea that a Sign determines an Interpretant in some ‘actual’ or ‘potential’ Mind (in other passages, a ‘quasi-mind’. See CP 4.536). On the grounds of this idea, we will differentiate between ‘potential’ and ‘effective’ semiosis. Potential semiosis is defined here as a triadically-structured process which is not taking place, which is only in potentiality. Effective semiosis, in turn, is a Sign in effective action, i.e., a Sign which, by being actualized, has an actual effect on the interpreter. Following the distinction between potential and effective semiosis, we can define potential information as a process of communicating a form which could be instantiated in a given moment, while effective information is the communication of a form from an Object to an Interpretant through the Sign, i.e., a Sign in effective action.
Chapter 4. Information and semiosis in Peirce’s science of signs 95
According to our interpretation of Peirce’s ideas quoted above, information has a processual nature: it is a process of communicating a form to the Interpretant which operates as a constraining influence on possible patterns of interpretative behavior. When applying this general semiotic approach to biological systems, information will most often be an interpreter-dependent process. It cannot be dissociated from the notion of a situated (and actively distributed) communicational agent. It is interpreter-dependent in the sense that information triadically connects representation (Sign), Object, and an effect (Interpretant) on the interpreter (which can be an organism or a part of an organism). The form — as a regularity embodied in the Object — acts as constraint on the interpreter’s behavior, but the interpreter always reconstructs the form (or at least some significant aspect of the form54) of the Object when interpreting a Sign. The interpreter does so such that an invariance is retained, which makes possible, in fact, the very act of interpretation. In sum, information in a biological system depends on both the interpreter and the Object (in which the form communicated in information is embodied as a constraining factor of the interpretative process). s a way of stressing the difference between this account and more usual explanations about what is information, consider, for instance, Maynard Smith and Szathmáry’s (1999: 9–10) argument that information is ‘that something’ which is conserved throughout a series of changes in the material medium underlying a communication process. We see this as resulting from a tendency to substantialize information. According to the account developed above, ‘that conserved something’ is not information, but rather an invariance in the reconstructed form. Information is rather the process by which a form is communicated through several different media (Signs) in such a way that an invariance is conserved throughout the process, even though the significant aspects of the Object’s form are continually reconstructed. A framework for thinking about information as a process can be built in Peircean terms by employing the following definitions:
54
The reader can relate here the notion of ‘form’ with the ‘ground’ of a Sign (see CP 2.228). Although they are not equivalent, they can be seen as overlapping in several respects. A discussion of this feature will be developed in future works.
96
Genes, Information, and Semiosis
[Information = semiosis] A triadic-dependent process through which a form embodied in the Object in a regular way is communicated to an Interpretant through the mediation of a Sign. [Potential information = potential semiosis] A process of communicating a form from an Object to an Interpretant through the mediation of a Sign that could take place in a given moment. [Effective information = effective semiosis] The process by which a Sign effectively produces an effect (Interpretant) on some system (an interpreter) by making the Interpretant stand in a similar relation to something else (the Object of the Sign) as that in which the Sign itself stand. Thus, the Sign mediates the relation between Object and Interpretant. The Sign effectively communicates, in this way, a form from the Object to the Interpretant, changing the state of the interpreter. To formulate the above definitions in a sufficiently clear way, we should define what we mean by ‘process’. We follow here Rescher in his definition of a process as “... a coordinated group of changes in the complexion of reality, an organized family of occurrences that are systematically linked to one another either causally or functionally” (Rescher 1996: 38). In the next chapter, we will address the relationship between the Peircean semiotic approach to information we developed here and some other accounts found in the vast literature on this issue.
Chapter 5 Some ideas concerning the relationship between the Peircean approach and other ideas about information
The Peircean semiotic approach to information we developed in the previous chapter raises several questions and shows both similarities and differences when contrasted with other accounts of information. We do not intend, however, to present in this book any exhaustive discussion about the relationships between our account and other approaches to information. In this chapter, we will only stress some similarities and differences between a Peircean semiotic treatment of information and some ideas put forward by Gregory Bateson and Eva Jablonka (see also Queiroz et al. 2008).
1. Information theory As discussed in Ch. 3, Shannon and Weaver’s mathematical theory of communication55 includes a measure of the amount of information in terms of the unexpectedness of a sequence of signals, written H=∑pilog(1/pi), where pi is the probability of the ith form of a signal. As it is well known, this probabilistic measure of information is nonsemantic, and, even though it can be useful in biological research for several purposes, it is not clear whether it can be enough for understanding biological information, and, moreover, there are arguments against this prospect (see Ch. 3). An important point to highlight, then, is that the Peircean account of information we developed shows an obvious difference from Shannon and Weaver’s approach, since it takes into account the semantic and pragmatic dimensions of information. The focus in a Peircean approach to biological information is naturally on the meaning of Signs to a given living system and the 55
The book by Shannon and Weaver from 1949, called The Mathematical Theory of Communication, became almost synonymous with ‘information theory’. For biological applications, see Yockey (1992), or the work of Schneider’s lab (e. g., Schneider, Stephens 1990; Schneider 1994). The theory developed in the 1949 book is mainly based upon Shannon’s (1948) paper. Other important ideas preceded it and other kinds of mathematical or algorithmic approaches to information have been developed later on. For a biological application of algorithmic information theory, see Küppers (1990). In the humanities, “communication theory” is sometimes referred to as denoting the interdisciplinary field of human communication, including pragmatics, sociolinguistics, rhetorics, etc.
100
Genes, Information, and Semiosis
variation shown by meanings in different contexts of interpretation. Thus, we can argue that a major advantage of this approach when applied to living systems, as compared to the mathematical theory of communication, is that it coins a semantic/ pragmatic concept of information, addressing an open problem in the philosophy of biology (e. g., Küppers 1990; Jablonka 2002, 2005; Roederer 2005). Obviously, this does not mean that Shannon and Weaver’s approach is not useful in the domain of biology; rather, it is clear that it brings contributions to the treatment of certain issues in which the meaning of Signs and the contexts of interpretation are not particularly relevant (see Yockey 1992; Maynard Smith 1999, 2000a,b; Adami 2004). In the case of other problems, in which the meaning of Signs is regarded as a fundamental feature of information systems, networks, pathways, another conceptual framework, which takes in due account semantics and pragmatics, is needed. This does not mean that the Peircean approach to signals, signs and the analysis of, e. g., sequence information in fields like molecular biology, bioinformatics and systems biology address a totally different set of problems. Rather, as we see it, the Peircean approach to what is ‘biologically meaningful’ in such fields has simply not yet been developed, and it is highly probable that further integration of such an approach with the more quantitative approaches in, for instance, bioinformatics could be developed in the future.
2. Bateson’s understanding of information as a difference which makes a difference As an alternative to the treatment of biological information as just sequence information in nucleic acids and polypeptides, it has been proposed to use Gregory Bateson’s (1972, 1979) understanding of information (e. g., Hoffmeyer, Emmeche 1991; Emmeche 1990; Hoffmeyer 1996, 2008b, Bruni 2003, 2007). Bateson (1972: 453) conceives ‘information’ (or, as he stressed, the ‘elementary unit of information’) as “a difference which makes a difference”. As we think that Bateson’s definition of information involves the idea that, for something to be ‘information’, it is necessary some interpreting system that, by interpreting it, suffers its effects, we would add that a
Chapter 5. Some ideas concerning the relationship between ...
101
difference can only make a difference to ‘somebody’, or, more generally speaking, to an interpretative system.56 Nevertheless, what is a difference? Bateson asks where are the differences between two objects and answers that they are not in any of the objects, nor in the space between them, nor in the time between them. He comes to the conclusion, then, that a difference is an ‘abstract matter’ (Bateson 1972: 452). Bateson stresses the contrast between the world as seen from the perspective of the physical sciences (which he calls ‘pleroma’), a world in which effects are caused by rather concrete conditions or events, impacts, forces, etc., and the ‘world of communication’ or ‘organization’, the ‘psychological’57 world (which he calls ‘creatura’), in which effects (if such a word can still be used) are brought about by differences (Bateson 1972: 452).58 In an effort to clarify the abstract concept of ‘difference’, Bateson argues that the word ‘idea’, in its most elementary sense, is synonymous with ‘difference’. In any thing, say, a piece of chalk, there are an ‘infinite’ number of ‘differences’ around and within it, differences between the chalk and any other thing in the universe, and, “within the piece of chalk, there is for every molecule an infinite number of differences between its location and the locations in which it might have been” (Bateson 1972: 453). It is precisely because of this infinitude, Bateson argues, that a piece of chalk, or any other thing, cannot enter into communication or mental processes as Ding an sich. He observes that we (or, generally speaking, any interpreter) ‘select’,
56 This is not the first time that Bateson is interpreted this way (see, e. g., Wilden 1980). For an explicit connection of the expression ‘makes a difference’ to the notion of ‘somebody’ or an organism, see also Emmeche (1990: 53; 2000) and Hoffmeyer (1998). 57 Bateson’s understanding of what is a ‘psychological’ world obviously depends on the broad concept of mind he assumes. It would be an interesting task to compare Bateson’s and Peirce’s concepts of ‘mind’, but we should leave it to future works. 58 Notice that we do not refer to the ‘physical world’ and ‘the mental world’, but, rather, to the world seen as ‘physical’ or as ‘mental’. This is indispensable, for reasons of consistency with Bateson’s notion of ‘mind’ and his general critique of the separation of mind and body. Symptomatically, he writes that the world in which effects are brought about by differences (‘creatura’) is “... the world seen as mind, wherever such a view is appropriate” (Bateson 1972: 457).
102
Genes, Information, and Semiosis
‘filter out’ a ‘very limited’ number of differences around and within the piece of chalk, which ‘become information’ (Bateson 1972: 453). We defined ‘information’ as a triadic-dependent process through which a form embodied in the Object in a regular way is communicated to an Interpretant through the mediation of a Sign (see Ch. 4). We can notice, then, an important difference between this Peircean account and Bateson’s treatment of information. When Bateson argues that information is a difference which makes a difference, he seems to be focusing on the form communicated (the ‘difference’), rather than on the process of communicating the form. The latter would be rather conceptualized, in Bateson’s account, as the process through which a difference (‘information’) ‘makes a difference’. Thus, even though both accounts can be seen as committed to an interpretation of information as having the nature of a process, this nature is, in our view, quite explicit highlighted in the Peircean definition of information proposed in the previous chapter. Nevertheless, we see these accounts as intersecting and not in conflict, even though they certainly diverge in important respects. In our view, Bateson’s notion can be interpreted as triadic-dependent and the important aspects of both notions are processual and relational. From the point of view of a Batesonian ‘ecology of mind’ (or, as some would prefer, ‘second-order cybernetics’59), a living system entails ongoing semiosis as the action of Signs, where any single Sign is a first that stands (by a code or a habit) in such a relation to a second, its Object, so as to determine a third, its Interpretant, to take the same relation to that Object (that the Sign takes) and thereby effecting that Interpretant so that this effect is significant (potentially or actual) to that Interpretant’s interpreter-organism, in the sense that it is a difference that makes a difference to the interpreter. The interpreter must in this Batesonian perspective be an organism-environment unit of survival (or a part of an organism, or an organism-like entity, within 59
The very fact that second order cybernetics emphasizes the role of the observer in investigating cybernetic systems makes this approach more prone to being integrated with a semiotic approach (see also Brier 1996). It is disputed whether there is a clear ‘break’ between first and second order cybernetics; thus, according to Heylighen, Joslyn (2001), “if we look more closely at the history of the field, we see a continuous development towards a stronger focus on autonomy and the role of the observer, rather than a clean break between generations or approaches”. Furthermore, “the second order perspective is now firmly ingrained in the foundations of cybernetics overall” (ibid.).
Chapter 5. Some ideas concerning the relationship between ...
103
such a unit), and the effects on that organism’s parts, to be significant (i.e., ‘to make a difference’), cannot be merely physical, because by definition, the difference, if any, that they make, is of potential or actual purport or relevance to the organism in question, which means that they concern the organism’s chances of finding food or other sources of energy, or that they ultimately concern its chances of surviving and reproducing. It is important to notice that we are not advocating here a synthesis of Peirce’s and Bateson’s framework in toto.60 Our main point of reference is the Peircean framework, and, from this standpoint, we think it is both useful and inspiring to look for points of convergence, similarity, compatibility, or even possibilities of synthesis of some notions in the works of authors otherwise different. Bateson’s and Peirce’s views are surely very different, but we don’t think it is not possible to pragmatically introduce some ideas from Bateson’s works within a Peircean framework.
3. Jablonka’s concept of semantic information Recently, Jablonka (2002) proposed a semantic definition of biological information. She suggested a list of requirements to identify a ‘common denominator’ among informational phenomena of different types (alarm calls, DNA sequences, pieces of software, etc). The types include ‘environmental cues’, ‘man-made instructions’, ‘evolved biological signals’, and ‘hereditary material’. The common attributes are: (i) a special type of reaction between receiver and source (potential or actual action of the receiver); (ii) that the receiver’s response depends on the organization of the source, i.e., the response is not dependent on matter and energy transference; (iii) that the reaction to the source contributes to the evolution of the receiver; (iv) that 60
On some of the ‘Gatherings in Biosemiotics’ meetings, there were debates about the possibility of ‘uniting’ the theories of Peirce and Bateson. Peirce-scholar Edwina Taborsky has been skeptical due to ‘nominalist’ tendencies in Bateson’s thinking, and Bateson-scholar Peter Harries-Jones has been skeptical due to the lack of eco-social elaborations in Peircean semiotics. We do not think that comparative studies of Bateson and Peirce has advanced far enough to allow forecasting the possibilities of any overall united cybersemiotic metaphysics, and such a project is not our aim here.
104
Genes, Information, and Semiosis
variation in the form of the source produces a corresponding variation in the form of the response. On the grounds of these attributes, Jablonka proposes a definition framed by a functional-evolutionary perspective that emphasizes the prominent rule of the interpretative system of the receiver in evolutionary terms. This definition brings, in her view, the following advantages: (i) it accommodates environmental cues and potential informational sources; (ii) it makes it easy to think about non-genetic information. The definition is as follows (Jablonka 2002: 582): A source — an entity or a process — can be said to have information when a receiver system reacts to this source in a special way. The reaction of the receiver to the source has to be such that the reaction can actually or potentially change the state of the receiver in a (usually) functional manner. Moreover, there must be a consistent relation between variations in the form of the source and the corresponding changes in the receiver.
According to this model, a source has information when the modification of a receiver system is functionally coupled to the variation of the source of the form. But for a source to be regarded as informational, it must elicit an adaptive response — a functional reaction in an evolutionary sense. Jablonka (2002: 582) explains that ‘functional’ is used in her definition to mean the “consistent causal role that a part plays within an encompassing man-designed or natural-selectiondesigned system, a role that usually contributes to the goal-oriented behavior of this system”. This is related to her definition of the function of a part or a process as something that [...] has to be analyzed in terms of its causal role in the receiver system, which now or in the past contributed to the designed (by natural selection or by human intelligence) goal-oriented behavior of the encompassing whole. (Jablonka 2002: 584)
She argues that this is a combination of the notion of function as conceived by Cummins (1975), who defined the function of a given item P in a system in terms of the causal role P plays in the encompassing system’s behavior, with Wright’s (1973) definition of
Chapter 5. Some ideas concerning the relationship between ...
105
the function of a part of a system as the part’s effects which have evolved through natural selection.61 It is worth presenting some further definitions of terms put forward by Jablonka. She explains ‘form’ as the organization of a source’s features or actions, in particular those related to the actual or potential responsiveness of a receiver. The input, or information cue, is the source eliciting a specific, functional and regular response of the receiver. A signal, in turn, is an evolved informational input, and, finally, a code is a set of constraints establishing the possible adaptive coupling between signal and interpreter. We can clearly map semiotic concepts onto Jablonka’s definition: A ‘source’ can be defined in Peircean terms as an Object. A receiver system, in turn, can be understood as an interpreter, which has, according to Jablonka, an interpretative system that plays, she argues, a central role. She compares her definition of biological information with Maynard Smith’s (2000), claiming that, although both are based on evolutionary considerations, Maynard Smith requires that both the input and the final response (output) must have evolved by natural selection, while she requires instead that the interpretation and evaluation processes of the receiver are products of natural selection (Jablonka 2002: 582). Furthermore, these processes develop in a context-sensitive manner. As a consequence, her definition does not require that the form of an input evolves through natural selection; for instance, the responsiveness of a given living system to black clouds is a product of natural selection, but not the ‘blackness’ of the clouds. The perceptual and cognitive processes of the receiver evolved to be responsive to the form of the source’s variation. According to Peirce’s model, the [0]‘Blackness’ embodied in the Object is a necessary 61
The compatibility between Cummins’ and Wright’s approaches is a controversy matter in the literature and it does not seem clear that the synthesis suggested by Jablonka could be accepted without reservations. Godfrey-Smith (1998), for instance, considers that there is no obstacle to accepting both Cummins’ and Wright’s analyses of function as legitimate, although it is commonly held in the philosophical literature that these two views are competing analyses of the same concept. In his view, this shows that ‘function’ is an ambiguous term in scientific contexts, and even though he identifies a common core in the two concepts of function, he does not go as far as suggesting a synthesis, as Jablonka does. Recent criticisms raised by Cummins (2002) against Wright’s approach to function are telling with regard to the difficulty of proposing such a synthesis. It is beyond the scope of this book, however, to enter in more details about this issue.
106
Genes, Information, and Semiosis
requisite for the selection (in terms of semiotic analysis), and a regular spatiotemporal correlation ‘blackness — rain’, a necessary requisite for the interpretative process. The regularity of this spatiotemporal coincidence (‘blackness — rain’) is the form communicated from the source to the interpreter eliciting a specific adaptive response: “In order for external, non-evolved cues like a black cloudy sky be interpreted adaptively, the interpretation system of the receiver must be able to respond to the cloudy sky, a recurrent environmental agent, by specifically altering its internal state” (Jablonka 2002: 583). That is, a quality (e. g., blackness) embodied in the Object in a regular way is a requisite for the interpretative process. The emphasis on ‘form variation’ has interesting consequences when compared to Bateson’s interpretation of information as ‘a difference that makes a difference’ to somebody. As we saw above, Jablonka explains the ‘form’ of the source as the organization of the features and/or actions of the source, focusing specifically on those aspects of the organization to which the receiver reacts in a (usually) functional way. This is consistent, in our view, with the conceptualization of information as a process through which a ‘form’ is communicated from the Object to the Interpretant through the Sign. But, then, it would be necessary to avoid arguing, as Jablonka does, that an entity or a process can ‘have information’. Rather, an entity or process is said, in the account proposed in this book, to possess a ‘form’, to which an interpreter reacts when a Sign mediates a relationship between that entity or process and an effect on the interpreter. Therefore, a consistent relationship between variations in the form of the Object and the corresponding effects on the interpreter (Interpretants) results from the mediation of a Sign. However, the functional role of the Sign is not explicitly articulated in Jablonka’s account. The relationship between variations in the form of a process or entity and the corresponding effects on an interpreter is crucial in Jablonka’s account. The specific ‘reaction’ of a receiver system to the source (or, in more precise terms, the effect of the source on the interpreter) corresponds, in Peircean terms, to the Interpretant. In a way reminiscent of Bateson’s distinction between creatura and pleroma, Jablonka argues that, for a source to be an information input rather than a source of energy or material, its form or variations in its form, rather than any other attribute, should affect the interpreter’s
Chapter 5. Some ideas concerning the relationship between ...
107
response in a consistent, regular way (Jablonka 2002: 585). That is, only when an entity or process is a difference which makes a difference to an interpreter, we can argue that information enters the scene. The relationship between the form of the Object and the Interpretant (in Peircean terms) involves in living systems complex and regulated chains of events, and, as Jablonka stresses, “in all cases this chain of events depends on the way the source is organized rather than on its energy content or its precise chemical constitution” (Jablonka 2002: 580).62 Jablonka’s definition of the form of a source can be related to Peirce’s notion of form as a habit or a rule of action. Nevertheless, when we compare Jablonka’s account of information with the Peircean approach developed here, we can detect some important points of disagreement. Her model, as described in the 2002 paper, seems to be dyadic, i.e., she seems to lose from sight the Sign as the agent mediating the relation between Object and Interpretant, and, consequently, she does not explicitly recognize that the form of the Object is communicated to the interpreter through the mediation of the Sign. While analyzing Jablonka’s definition of information, we should ask how the form of the source is communicated to the interpreter. In the Peircean approach to information developed here, the answer is that Signs mediate the relation between Objects and Interpretants, and, thus, bring about a consistent relation between variations in the form of the Object and corresponding effects on the interpreter (Interpretants), and that this can happen in many different ways, depending on the types of Signs, Objects and Interpretants involved. When Jablonka argues for the generality of her definition, as applying to all types of information, she writes (Jablonka 2002: 585): [...] a source S (allele, alarm call, cloudy sky, etc.) carries information about a state E for a receiver R (an organism or organism-based product), if the receiver has an interpretation system that reacts to S in a way that usually ends up adapting R (or its designer, if R is humanly designed) to E.
62
For interesting comparisons between informational processes, dependent on the organization of the source (Object), and non-informational processes, involving material and energy transfer, and, accordingly, dependent on energy content or chemical constitution, and also between sources of information and sources of material and energy, see the original paper.
108
Genes, Information, and Semiosis
A comparison between the definition she offers in page 582 of her paper (Jablonka 2002) and this latter passage shows how a semiotic treatment can bring important contributions to Jablonka’s approach. Mapping semiotic concepts onto the latter passage, we obtain a picture which is significantly different from that resulting from the previous quotation: As a ‘source’ is now explained as ‘carrying the information’, it might be defined in Peircean terms as a Sign, introducing the missing element in her definition. The Object is, in the latter case, a state E to which the receiver is adapted. As above, the receiver’s interpretation system is the interpreter, and the ‘reaction’ of a receiver system to the source, the Interpretant. Finally, in the conclusions of her paper, Jablonka (2002: 602) writes: […] a source becomes an informational input when an interpreting receiver can react to the form of the source (and variations in this form) in a functional manner.
In this statement, it seems that Object and Sign are differentiated, through the usage of the concept of ‘input’, which may be read in Peircean terms as a Sign, while the source would be the Object. Furthermore, the same element first plays the role of Object (source) and subsequently of Sign (informational input) when the interpreter enters the scene. Jablonka’s scheme is entirely interpreter-dependent, as she herself emphasizes (Jablonka 2002: 582), but a crucial idea in a Peircean framework is not clear in it, namely, that the semiotic process is irreducibly triadic, i.e., the three elements of a semiotic process are necessarily and all the time interdependent. In other words, a Peircean account leads to an emphasis on the very prominent role of the interpretative system of the receiver. But all the process (cf. triadically interdependent model) is highly distributed, and there shouldn’t be any prominence of a special component. In a Peircean model, Sign, Object and Interpretant are triadically coupled in a dynamic irreducible process. In other words, information requires a triadic pattern of determinative relationships involving S-O-I. In our view, the usage of Peirce’s conceptual toolbox would help avoiding the vacillation we observe in the way Jablonka explains the elements in her concept of information, with the Sign being sometimes left outside the picture, and, when introduced, being sometimes conflated or even merged with the Object. In our view, these are
Chapter 5. Some ideas concerning the relationship between ...
109
consequences of the lack of a Sign-theoretical framework in her account of information in biology. Another potential problem in Jablonka’s account lies in her claim that ‘information’ is conferred by the receiver (Jablonka 2002: 586). After all, there is a real aspect in the environment which is necessary for information to take place. This is yet another point in which Peirce’s account is helpful. It is an important assumption in a Peircean framework that the ‘blackness’ of a cloud, for instance, is a form embodied in an Object in a regular way. The regular property of blackness and the ‘blackness — rain’ coincidence compose the form communicated from the source to the interpreter eliciting a specific response. In this case, we treat information as the communication of a regular spatiotemporal correlation ‘blackness — rain’ from O to I. The communication of a form is the transference of this correlation to the interpreter so as to produce a specific response, an effect on the interpreter, constraining its behavior. This brings about a constrained set of effects of the Object on the interpreter through the mediation of the Sign.
4. Concluding remarks Both Bateson’s and Jablonka’s concepts of information (or, for that matter, the Peircean account developed in this book) may seem too broad, but we think they are as broad as they should be, since information is itself a sweeping concept. Information can encompass a variety of processes, involving, for instance, genes, molecules, computers, the media, and everyday things such as recipes or instructions in a manual. Furthermore, information can be acquired, communicated, reconstructed, processed, translated, shared, etc. in a variety of ways. Therefore, we can say that, in an adequate manner, these accounts are as broad as the phenomenon they intend to grasp: A wide variety of entities and processes can be differences which make a difference to an interpreter or can make a receiver system reacts to them in such a way that the reaction can actually or potentially change the state of the receiver in a (usually) functional manner, or can embody a regular form which may be communicated to an Interpretant through the mediation of a Sign.
110
Genes, Information, and Semiosis
Symptomatically, Jablonka (2002) presents arguments to the effect that one of the advantages of her definition is its broad nature, allowing it to accommodate information as related to both environmental cues and evolved signals. Furthermore, she argues that the definition can be used as a basis for a comparative analysis of different types of information systems in living beings. It also avoids an attribution of a theoretically privileged informational status to genes, which are just one of the types of informational sources contributing to the development and functioning of organisms. Another important aspect to be noted is that she restricts information to living systems (and systems designed by living systems, such as man-made devices): “... according to my definition, information is something that can exist only when there are living (or more generally, designed) systems. Only living systems make a source into an informational input” (Jablonka 2002: 588). We are also sympathetic to the idea that living systems have been the first genuine semiotic systems.63 In our view, these systems have been, particularly, the first cell-like entities, in which there was a boundary separating an internal from an external environment, thus requiring that the system interpreted external entities and processes as meaning something more than just being external events, as a part of cosmos with no pragmatic significance, but rather being potentially useful (or the opposite) for the maintenance and reproduction of the system, i.e., being relevant Signs, and, furthermore, internalized such a meaning, producing another Sign inside the system, which stood for the Object as the external Sign itself stood, i.e., an Interpretant (cf. Hoffmeyer 1998). Bateson’s and Jablonka’s remarks on information also suggest — as our account — a process interpretation of the concept. Even though it may not seem so clear in Bateson that information is a process, his arguments indeed focus on a dynamical process by which a difference makes a difference to a system which interprets it. Similarly, Jablonka does not make it clear that information can be conceived as a process, but she stresses that the source is made into an information input when the receiver functionally reacts to it, and this highlights, in turn, the 63 We will not enter into the debate about borderline cases or the famous question of the semiotic threshold below which processes cannot be characterized as depending on triadic relations or Thirdness. On this issue, see Nöth (2001) (and the special issue of Sign Systems Studies, vol. 29, issue 1, 2001, as a whole), and Stjernfelt (2006).
Chapter 5. Some ideas concerning the relationship between ...
111
process mediated by the interpretative systems of living beings which functionally correlate the variation of the form of the source to the variation in the response of the receiver. Our account, in turn, makes it explicit the conceptualization of information as a process, namely, the process of communicating a form from the Object to the Interpretant through the Sign. In short, we believe it is possible to employ the account of information developed in the previous chapter as a basis for building a synthetic account, incorporating several aspects of both Bateson’s and Jablonka’s accounts. We hope this challenge will be taken up in future works.
Chapter 6 A semiotic analysis of genes and genetic information: first take
1. Information talk in action: a textbook description of the ‘flow of information’ in a cell The analysis we will perform here requires that we first present some very general notions about key processes in the genetic information system, namely, transcription, mRNA splicing, and protein synthesis. This section intends to present these basic notions, but it also plays another role in our arguments: we will use it to show how information talk pervades descriptions and explanations in molecular biology and genetics. For these purposes, we present a typical description of the ‘flow of information’ in a cell, according to the central dogma of molecular biology. The section has been built around the treatment of this topic in a widely used textbook, Benjamin Lewin’s Genes VIII, but other textbooks have also been used as sources (Griffiths et al. 1999; Lodish et al. 2003; Alberts et al. 2002; Klug, Cummings 2002). We deliberately avoided introducing a large number of details, which can be easily found in any molecular and cell biology textbook. First, we need to explain why we are using textbook science. As Kuhn ([1962]1996: 136) writes, science textbooks present an already articulated body of problems, data, and theory, i.e., a particular set of paradigms to which a scientific community is committed. Or, to put it differently, textbooks explain the bases of the current normal-scientific tradition, and can be seen, thus, as “pedagogic vehicles for the perpetuation of normal science” (Kuhn [1962]1996: 137). In particular, they communicate the theoretical vocabulary and conceptual framework of a contemporary scientific language. Therefore, we can employ textbooks to show how information talk is a quite important element of normal science vocabulary in genetics and molecular biology, and, moreover, that precisely for this reason it should be properly explained (since we do not accept the idea that it should be simply eliminated). The ‘central dogma of molecular biology’ is typically conceived as a statement about the ‘flow’ of ‘information’ in a cell. A very simple model of the process of gene expression (Figure 3) shows the direction of this ‘flow’, with the exception of reverse transcription (not shown in the picture). Lewin describes gene expression as follows:
116
Genes, Information, and Semiosis
“the expression of cellular genetic information usually is unidirectional. Transcription of DNA generates RNA molecules that can be used further only to generate protein sequences; generally they cannot be retrieved for use as genetic information. Translation of RNA into protein is always irreversible” (Lewin 2004: 28. Italics indicate emphases in the original. Bold, emphases added). It is quite interesting to observe the usage of the metaphors of ‘transcription’ when one is dealing with the same ‘language’, the ‘nucleotide language’ of nucleic acids, and translation, when one considers two ‘languages’, as proteins are ‘written’ in an ‘amino acid language’. While explaining that a gene codes for a single polypeptide, Lewin describes how a gene ‘codes for’ an RNA, which may in turn ‘codes for’ a protein. This shows, in turn, why regions in DNA are said to ‘constitute genetic information’ (Lewin 2004: 22). He also states that “the basic principle is that the gene is a sequence of DNA that specifies the sequence of an independent product”, which is either an RNA or a protein (Lewin 2004: 16). The close connections between the central dogma and the classical molecular gene concept, discussed in Ch. 2, are quite clear in his arguments. During the synthesis of pre-mRNA, the four-base ‘language’ of DNA (as a sequence of nucleotides including the bases adenine, A, guanine, G, cytosine, C, and thymine, T) is copied or ‘transcribed’ into the four-base ‘language’ of RNA (with uracil, U, replacing T). Transcription results in functional mRNAs (messenger RNAs), rRNAs (ribosomal RNAs), tRNAs (transfer RNAs), snRNAs (small nuclear RNAs), and scRNAs (small cytoplasmic RNAs), but we will focus here on the synthesis of mRNA. Other functional RNAs which play important roles in various steps in DNA processing will be mentioned in passing. During transcription, one DNA strand acts as a ‘template’, determining by base pairing the order in which monomers (ribonucleoside triphosphates) are assembled to form a complementary RNA polymer, by a polymerization reaction catalyzed by the enzyme RNA polymerase. The effects of a protein-coding gene on a given cell or organism are regulated mainly by control of gene expression at the level of transcription initiation. The transcription of a gene can be either repressed, when the corresponding mRNA and encoded protein or proteins are synthesized at low rates or not synthesized at all, or
Chapter 6. A semiotic analysis of genes and genetic information
117
activated, when both the mRNA and encoded protein or proteins are, ceteris paribus, produced at much higher rates. Through the control of gene expression, only a subset of all genes present in any cell type in a multicellular organism is really expressed. Thus, from all the potential protein products a given cell type might have, only a specific number and variety will be present. This is the fundamental basis for cell differentiation in multicellular organisms.
Figure 3. A very simple model of the steps of gene expression. (From Campbell, Reece 2002.)
After addressing transcription, Lewin considers the issue that in more complex eukaryotic genes the immediate ‘transcript’ of a gene is a
118
Genes, Information, and Semiosis
pre-mRNA, which should be ‘processed’ in order to generate the mature mRNA, primarily involving a process known as RNA splicing. As we will systematically use as an example a gene which undergoes alternative RNA splicing, we will expand a little more on this process. As discussed in Ch. 2, by the end of the 1970s it was found that eukaryotic genes are split into exons, pieces of coding sequence, separated by introns, non-coding segments, which are excised from a ‘primary transcript’ or pre-mRNA, in a process known as RNA ‘processing’, which includes other events not described here. After the introns are excised, the coding exons are joined back together into a functional mRNA, which will be transported to the cytoplasm of the eukaryotic cell, where protein synthesis will take place. In Ch. 2, we also addressed alternative RNA splicing, an important mechanism for the production of different forms of proteins (isoforms) by different cell types. We presented as an example the fibronectin (FN) gene, which generates more than 20 different FN isoforms. The FN gene has approximately 75,000 nucleotides (75Kb) and contains numerous exons (Figure 4). After the FN pre-mRNA is transcribed from DNA, it undergoes cell type-, development- and agespecific splicing, acting as a template for the synthesis of specific FN isoforms, which result from a unique combination of exons found in the FN gene.
Figure 4. Cell type-specific splicing of fibronectin pre-mRNA in fibroblasts and hepatocytes. The 75-kb FN gene (top) contains multiple exons. Introns are shown in the diagram as thin lines and are not drawn to scale. Most of the introns are much longer than any of the exons. The FN mRNA produced in fibroblasts includes the EIIIA and EIIIB exons, whereas these exons are spliced out of FN mRNA in hepatocytes (modified from Lodish et al. 2000.
Chapter 6. A semiotic analysis of genes and genetic information
119
The combinations of exons in each isoform change its causal dispositions. This can be clearly seen in the case of the splicing of FN pre-mRNA in fibroblasts and hepatocytes. In fibroblasts, splicing of the FN pre-mRNA results in mRNAs containing exons EIIIA and EIIIB. The fibroblast FN isoform contains amino acid sequences that bind tightly to proteins in the plasma membrane, ascribing it specific causal dispositions. This specific FN isoform contributes to the adhesion of fibroblasts to the extracellular matrix. In hepatocytes, the major cell type in the liver, cell-type specific splicing results in functional FN mRNAs lacking exons EIIIA and EIIIB. As in the case of fibroblasts, we have here a FN isoform with specific causal dispositions. First, it does not show the causal dispositions of the fibroblast isoform: FN secreted by hepatocytes does not adhere tightly to fibroblasts or most other cell types. The lack of such causal dispositions is very important to the functionality of this FN isoform, as it allows it to freely circulate in the blood stream. Nevertheless, when the wall of a vase is ruptured, hepatocyte FN plays a fundamental role in the formation of blood clots, showing its specific causal disposition, which result from the presence in the protein of fibrinbinding domains, amino acid sequences that bind to fibrin, one of the main constituents of blood clots. When hepatocyte FN is bound to fibrin, it shows yet another causal disposition, interacting with integrins, cell-adhesion protein molecules found in the membranes of activated platelets. As a result, the blood clot is expanded through the addition of platelets. The effects of genes on the functioning of a cell or organism can also be regulated by means of alternative pre-mRNA splicing, so as to produce different gene products from the same pre-mRNA. Particularly remarkable examples of genetic regulation at the level of RNA splicing are found, for instance, in the sex determination pathway of Drosophila (for a review, see, e. g., Black 2003). Finally, translation is an essential part of protein synthesis, consisting in the process by which the nucleotide sequence of an mRNA serves as a template for the synthesis of a polypeptide chain, i.e., for a series of events in which amino acids are ordered and joined to form the primary structure of a protein. Three types of RNA molecules are involved in translation, performing different but cooperative functions. mRNAs are the ‘vehicles’ of the genetic information transcribed from DNA. The ‘message’ at stake is ‘written’
120
Genes, Information, and Semiosis
in the form of a series of three-nucleotide sequences, called ‘codons’, each of which specifying a particular amino acid. tRNAs play a fundamental role in the process of deciphering the codons in mRNA. Each type of amino acid has its own subset of tRNAs. They act as transporters or adaptors, binding amino acids and carrying them to the growing end of a polypeptide chain in response to specific codons in the mRNA. The reason why the correct tRNA with its attached amino acid is selected at each step in protein synthesis lies in the fact that each specific tRNA molecule contains a three-nucleotide sequence, called an ‘anticodon’, that base-pairs with its complementary codon in the mRNA. In this manner, for each specific codon in mRNA a specific amino acid, carried by a specific tRNA, is included in a polypeptide chain, according to the rules expressed in the almost universal ‘genetic code’. Along with 100 different proteins, several types of rRNA are components of ribosomes, the complex and large macromolecular structures that act, so as to say, as guides to coordinate the assembly of the amino acid chain of a protein. In fact, an rRNA (a ribozyme), and not a protein, is probably the catalyst involved in the formation of peptide bonds in protein synthesis. Furthermore, the major movements of the ribosome are due to rRNAs. Translation involves three stages: initiation, when ribosomal units assemble near the translation start site in the mRNA with the tRNA carrying the amino acid methionine base-paired with the start codon, most commonly AUG; chain elongation, in which a four-step cycle is repeated, involving the binding of a tRNA carrying an amino acid, the release of the tRNA involved in the previous step in the elongation, transfer of the growing polypeptide to the incoming amino acid catalyzed by one of the rRNAs, and translocation of the ribosome to the next codon in the mRNA; and, finally, termination of synthesis, in response to stop codons UAA, UGA, and UAG. Recognition of a codon in mRNA specifying a given amino acid by a particular tRNA is, in fact, the second step in ‘decoding’ the genetic ‘message’. The first step is the attachment of the appropriate amino acid to a tRNA in a reaction catalyzed by a specific aminoacyltRNA synthetase. The specificity of the attachment between amino acids and tRNAs results from the capacity of each one of these enzymes of recognizing one amino acid and all its compatible, or ‘cognate’, tRNAs. Therefore, the rules captured in the genetic code
Chapter 6. A semiotic analysis of genes and genetic information
121
ultimately depend on the recognition activity64 of aminoacyl-tRNA synthetases. Although the terms ‘translation’ and ‘protein synthesis’ are usually employed interchangeably, this is not correct, since, although translation is obviously an essential step in protein synthesis, this process involves further steps, all of which can undergo regulation. Polypeptide chains undergo post-translational folding and often other changes, as, for instance, chemical modifications and association with other polypeptide chains, which are required for production of functional proteins. Lewin writes that, in the case of a protein that acquires its mature conformation spontaneously, i.e., that is capable of self-assembling, its three-dimensional structure simply follows from its primary sequence of amino acids (Lewin 2004: 22): By determining the sequence of amino acids in each protein, the gene is able to carry all the information needed to specify an active polypeptide chain. In this way, a single type of structure — the gene — is able to represent itself in innumerable polypeptide forms.
There are a number of proteins, however, that cannot self-assemble and should be assisted by other proteins, called chaperones, in order to acquire their proper structures. In this case, we might say that the gene does not ‘carry’ all the ‘information’ needed to specify an active polypeptide chain. Nevertheless, protein folding, generally speaking, is not treated by some researchers in the fields of genetics and molecular biology as such a simple and straightforward step that DNA can be said to encode all the information needed for it. Guimarães and Moreira (2000), for instance, argue that the expression of gene products always involves indetermination, since it is affected by several environmental factors, both internal and external to the cell. This is one of the reasons why ‘determination’ is, in their view, a term which should be applied in biology in a very specific sense, and, in the case of protein synthesis, is limited to the primary structure of polypeptides, i.e., merely the sequence of amino acids in the linear backbone of the polypeptide is specified by the sequence of 64
Molecular recognition is a very general phenomenon in molecular biology, involved in all enzyme-substrate interactions, and can also be described in its semiotic aspects (see Stjernfelt 1992).
122
Genes, Information, and Semiosis
nucleotides in DNA, through an mRNA intermediate, and with occasional mistakes. That a gene product is directly and immediately functional, only through translation of the mRNA sequence, is less common than textbook descriptions suggest. More frequently, protein function is achieved subsequently, as the activity of a protein usually depends on several genes, post-translational modifications and/or its integration in complex multimolecular systems. Protein ‘self-assembly’ is restricted, therefore, not only by the action of chaperones, but also by several other features of the processes which lead to a functional protein. As Guimarães and Moreira (2000) sum up, DNA codes for ‘functional possibilities’, rather than ‘functional products’. As discussed above, the genetic code is the short-hand expression in molecular biology for the set of rules according to which codons in mRNA are translated into amino acids during protein synthesis (Figure 5), thanks to the specificity of aminoacyl-tRNA synthetases. While addressing the genetic code, Lewin writes that “each gene represents a particular protein chain”, consisting of a particular series of amino acids. This means that we should face “[…] the issue of how a sequence of nucleotides in DNA represents a sequence of amino acids in protein” (Lewin 2004: 21. Emphases added). Or, to put it differently, the sequence of nucleotides in DNA “[…] codes for the sequence of amino acids that constitute the corresponding polypeptide” (Lewin 2004: 21, emphasis added). He characterizes, then, the genetic ‘code’ as the relationship between a sequence of nucleotides in DNA and a sequence of amino acids in the corresponding protein. He goes on to explain that “the genetic code is deciphered by a complex apparatus that interprets the nucleic acid sequence”, emphasizing that “this apparatus is essential if the information carried in DNA is to have any meaning” (Lewin 2004: 22. Emphases added). A codon is defined by him as a trinucleotide sequence “[…] representing one amino acid”. This leads to a description of a gene as including “[…] a series of codons that is read sequentially from a starting point at one end to a termination point at the other end” (Lewin 2004: 22, emphases added). He describes, then, how [...] there are three possible ways of translating any nucleotide sequence into protein”, namely, the ‘reading frames’ of a nucleotide sequence. Finally, he writes that in ‘translation’, tRNAs are involved in the process of ‘recognizing’ which amino acid corresponds to a particular nucleotide triplet in an open ‘reading frame’. (Lewin 2004: 26)
Chapter 6. A semiotic analysis of genes and genetic information
123
Figure 5. The genetic code. Sets of three nucleotides (codons) in an mRNA molecule are translated into amino acids during protein synthesis according to the rules shown in the table above. (From Griffiths et al. 1999.)
We can now invite the reader to check the terms highlighted in the previous paragraphs (either by italics or commas). They bring to light the conceptual problems one has to solve in order to ascribe a precise meaning to ‘information’ and the plethora of associated notions which compose information talk. What does it mean to say that genes code for proteins? Or that genes represent protein chains? Or that genes are structures that are able to represent themselves in polypeptide forms? Or that genes carry genetic information? Or that the genetic code is deciphered by an apparatus that interprets nucleic acid sequences and is essential if the information carried in DNA is to have any meaning? And so on. Those ideas and many other that appear in the passages quoted above should be properly explained, if we want to employ them in a manner which goes beyond mere metaphorical usage. In this book, we employ C. S. Peirce’s science of Signs to address some of these problems in a biosemiotic perspective, and, thus, fill in this conceptual
124
Genes, Information, and Semiosis
gap in genetics and molecular biology.65 It is time now to present the results we achieved in our semiotic analysis of the genetic information system. We will make it in two steps. First, we will put forward a general semiotic analysis of the gene as a Sign in this chapter. Then, in subsequent chapters, we will refine and extend this analysis.
2. A general semiotic analysis of the gene as a Sign If we take Peirce’s concepts of Sign and semiosis as bases for analyzing what is a gene, it will be the case that the action of a gene as a Sign will have to be understood as a relationship between three elements (Figure 6). Given the definition of information proposed in Ch. 4, genetic information can be described as a semiotic process. In these terms, there is clearly more to genetic information than just the sequence of nucleotides in a piece of DNA. In this picture, a string of DNA is a Sign. In this sense, the FN gene can be treated as a Sign. As a protein-coding gene, it stands — in a triadic-dependent relation — for a specific sequence of amino acids (Immediate Object) — one of the FN isoforms, translated out of a mature mRNA after alternative splicing (which can take place or not)66 — through a process of reconstruction of a specific form (Interpretant).67 65
As mentioned above, there are other versions of biosemiotics which do not take departure in the semiotics of Peirce, e. g., Barbieri’s (2003). Such versions may be in partial conceptual conflict and it may be also difficult to translate the one account of the code-aspect of the process of protein synthesis to the other account. Barbieri writes “A code […] requires three entities: two independent worlds and a code-maker which belongs to a third world” (p. 5); “in case of the genetic code, the codemaker is the ribonucleoprotein system of the cell, a system that operates as a true third party between genes and proteins” (ibid.). Barbieri links a notion of ‘ribotype’ (in his terms “the seat of genetic coding”, p. 157) to code-maker as a distinct category on par with phenotype and genotype. Barbieri’s notion of a code-maker as constituting an independent ‘world’ is not easy to translate to the semiotic notions used in this book. However, both approaches clearly locate one important argument for a semiotic or ‘semantic’ biology in the analysis of the genetic code. 66 If alternative splicing does not occur, it will be the case that Signs in DNA and Signs in mature mRNA will be equivalent. 67 In the case of genes, the Objects at stake are entities, as described above. Nevertheless, it is important to bear in mind that, in Peirce’s framework, it is not
Chapter 6. A semiotic analysis of genes and genetic information
125
Figure 6. A general semiotic analysis of the gene as a Sign.
A Sign is the mediating element in a semiotic process through which a form is communicated from an Object to an Interpretant. This is the reason why we consider the Interpretant here as the reconstruction of a form (habit) which was embodied in an Object. To be more explicit, we defined above information as the communication of a form from the Object to the Interpretant, and we also argued that such a communication will in turn constrain the behavior of the interpreter. What we mean by ‘reconstruction’ here is a process by which an aspect of the form68 of a protein (as a habit, a regularity) in a cell the case that the Object of a Sign should necessarily be an entity, a thing, or even an existent. Consider, for instance, the following passage: ‘The Objects — for a Sign may have any number of them — may each be a single known existing thing or thing believed formerly to have existed or expected to exist, or a collection of such things, or a known quality or relation or fact, which single Object may be a collection, or whole of parts, or it may have some other mode of being, such as some act permitted whose being does not prevent its negation from being equally permitted, or something of a general nature desired, required, or invariably found under certain general circumstances’ (CP 2.232). 68 Actually, it is not the complete form but only an aspect of a form of the protein that is communicated in this way, namely its primary structure (the sequence of amino acids), and this form or pattern in turn co-determines (eventually in
126
Genes, Information, and Semiosis
generation is communicated through Signs in DNA (in potency) to the form of a protein in the next cell generation, and the latter constrains the behavior of the cell as an interpreter. Thus, a regularity obtains (with obvious evolutionary consequences) in the three-dimensional structure and function of proteins over generations. We will introduce the qualifiers ‘Composite’ and ‘Simple’ to incorporate a part-whole relationship in our analysis, referring to a stretch of DNA or mature RNA as a whole as a Composite Sign, formed by clusters of Simple Signs, codons. We can now turn to a first refinement in our analysis, using the distinction between Immediate and Dynamical Object, and Immediate and Dynamical Interpretant in a systematic way (see Ch. 4). For an overview of the terminology, see also Table 1. The Dynamical Object of a gene is a functional, folded, and chemically modified protein, which is often not entirely specified by the sequences of nucleotides or amino acids, but it is rather indicated by such sequences. Functional proteins are not simply translated out of nucleotide sequences by a cell; they are rather found out through resources the cell acquire by collateral experience, i.e., by habits that a cell acquire in its development towards the states characteristic of a given cell type, and can be traced back to evolutionary processes.69 A functional FN isoform, for instance, is a Dynamical Object. interaction with other proteins, chaperones, as mentioned previously) the tertiary structure, i.e., the three-dimensional form of the protein synthesized. The gene as a Sign can only indicate, as it were, the protein; the cellular context co-determines the dynamical Interpretant of this Sign, as we shall see. This latter constraint is important to notice, since it shows that genetic preformationism (as if the gene as a Sign explicitly prescribed a protein structure in the next generation) does not follow from our analysis. See also the next two footnotes. 69 Symptomatically, Godfrey-Smith (1999) and Griffiths (2001) argue that developmental information is not stored in the genetic code, because the formal coding relation between codons in DNA and amino acids in polypeptides specifies only the primary structure of proteins. To be more precise, we should distinguish here between two situations: (a) Proteins that acquire their mature conformation spontaneously, ‘simply’ by self-assembly as related to their thermodynamic properties (yet to predict the 3D structure from primary structure is not so simple!). In this case, the three-dimensional structure of a protein seems to follow from its primary sequence of amino acids, and, therefore, the Immediate Object seems to directly determine the Dynamical Object (but notice the discussion about protein folding in the previous section). (b) proteins that do not self-assemble and should be assisted by chaperones in order to acquire their 3D structures. In this case, the
Chapter 6. A semiotic analysis of genes and genetic information
127
The Composite Immediate Object of a protein-coding gene is the sequence of amino acids of a polypeptide, since this is the object represented in the gene’s vehicle, a string of DNA.70 Each amino acid, sequence of amino acids, the Immediate Object, only indicates the functional protein, the Dynamical Object. In the text, we are dealing particularly with this case, which fits quite well Peirce’s understanding of the relationship between Immediate and Dynamical Objects. Chaperones can be treated, in these terms, as parts of the habits cells acquired through evolution. In both cases, however, the cellular context is crucial for determining the specific thermodynamic conditions for self-assembly (whether mediated or not by chaperones), and thus, as mentioned, in both cases, the “collateral experience” of the cell plays a role. 70 In our analysis, we are dealing with genes-D, as defined by Moss (2001, 2003), but by introducing the idea that the sequence of amino acids of a protein is the Dynamical Object represented in a Sign in DNA in its semiotically available form, we can be accused of introducing in our analysis a risky conflation between gene-P and gene-D. Even though we agree with Moss’ diagnosis that genetic determinism is supported by the confusion between genes as determinants of phenotypes (gene-P) and genes as developmental resources (gene-D), we think there is no problem with regarding genes as determinants of phenotypes at the level of the primary structure of a protein. Or, to put it differently, as regards the relationship between sequences of nucleotides and sequences of amino acids at the level of translation, we believe no serious problem follows from understanding the primary structure of proteins as being represented in DNA. We don’t see how genetic determinism, the main threat Moss has in view when vigorously criticizing the conflation of gene-P and gene-D, might follow from the claim that components of the primary structure of proteins are semiotically available in DNA in the form of nucleotide sequences. This problem only appears, in our view, when one considers that phenotypic levels higher than that of the primary structure of proteins are determined by genes. Genetic determinism is also avoided in the context of our analysis if we take into account that, yet at the cell level, functional proteins as Dynamical Objects are not fully determined by DNA sequences, but only indicated by them and found by cells through habits established through evolution and development. Thus, it is not even the case that the three-dimensional structure of a protein is completely determined by, or even represented in, DNA in our picture. Furthermore, our analysis keeps genes on the same plane as other biomolecules involved in development, giving them no causal privilege. We work here with genes as developmental resources which represent a range of possible proteins (their Immediate Interpretant), which will in turn be embedded in complex causal structures, in which many other molecules play important causal roles. In fact, we don’t see a possible conflation of gene-P and gene-D as a sufficient basis for arguing that one should abandon such a basic idea in molecular biology as that of genetic coding. We consider that there is a fundamental difference between talking about genes-D as representing amino acid sequences and talking about genes-D as determinants of organismic phenotypes.
128
Genes, Information, and Semiosis
in turn, is a Simple Immediate Object. If we consider the sequence of amino acids of a specific FN isoform, we will say, in the terms of our analysis, that such a sequence is an Immediate Object of the FN gene. It is important to bear in mind, however, that it is an Immediate Object, not the Immediate Object. After all, the FN gene codes for more than 20 different FN isoforms, all of them being potential Immediate Objects of the FN gene as a Sign in DNA. The sequence of amino acids, the Composite Immediate Object, is the Dynamical Object in its semiotically available form. The sequence of amino acids of each FN isoform amounts to a specific protein coded — in its semiotically available form — in a mature RNA which is a result, after splicing, of a pre-mRNA transcribed from the FN gene. The Immediate Object, a sequence of amino acids, can indicate a range of possible functional proteins, Dynamical Objects, since a single amino acid sequence can be folded in different ways in different cellular contexts. But we should not lose from sight, however, that such an indication by the Immediate Object plays a fundamental role in the reconstruction of one Dynamical Object, since it is not the case that any three-dimensional protein can be produced from a given amino acid sequence.71 The Immediate Interpretant of a codon as a Simple Sign is the range of interpretability established by the rules of base pairing by which specific nucleotides in DNA determine specific nucleotides in mRNA, or the range of interpretability of three-nucleotide sequences in mature mRNA as established in the genetic code, the set of rules by means of This latter way of talking should be avoided, as Moss rightly argues, since it involves a commitment to a preformationist, determinist view of the relationship between genotype and phenotype. The former, however, is in fact implied by an idea which appears in Moss’ works themselves, namely, that a gene-D is a specific nucleic acid template. We think our analysis is compatible with the idea that genes-D contain ‘molecular template resources’ involved in the synthesis of ‘gene products’, as Moss argues (see, e. g., Moss 2001: 88). 71 It is not the case, in the Peircean framework, that the Immediate Object is a condition of possibility to the Dynamical Object. Nevertheless, in the case we are analyzing here a Dynamical Object of a given class is created (due to the habits embodied in the cell as an interpreting system) on the grounds of indications present in the Sign. A cell uses Signs in DNA as a basis for synthesizing a Dynamical Object sufficiently resembling a past Dynamical Object which does not exist anymore but resulted in successful, adaptive experiences. This is the reason why we claim that, in this case, the Immediate Object establishes conditions of possibility for the Dynamical Object.
Chapter 6. A semiotic analysis of genes and genetic information
129
which nucleotide sequences determine the addition of specific amino acids to a growing polypeptide chain. Symptomatically, ‘coding’ can be defined as a system of constraints which establishes a range of possible effects of a Sign (see Nöth 1995: 210–211). The Dynamical Interpretant of a codon as a Simple Sign amounts, then, to the realization of one of the rules of base pairing or of the genetic code. A Composite Sign in DNA determines a range of possible Composite Immediate Objects. It is true that there are cases in which a stretch of DNA codes for only one protein product. In this case, the Composite Sign in DNA determines only one Immediate Object. Nevertheless, in eukaryotic cells at least, most stretches of DNA codes for several distinct proteins, as in the case of the FN gene. Therefore, we can define the Immediate Interpretant of a Composite Sign as the range of interpretability72 of that Sign in DNA, i.e., as the fact that it can be interpreted so as to produce a range of potential Immediate Objects, potential sequences of amino acids that can be produced from that Sign in DNA. Alternative RNA splicing is understood, in these terms, as one of several processes that enrich the range of interpretability, the Immediate Interpretant, of a stretch of DNA. In the case of the FN gene, its Immediate Interpretant comprises more than 20 possible Composite Immediate Objects. This analysis is in accordance with the definition of a Sign as medium for communicating the form of an Object to an Interpretant. The Interpretant can be seen, thus, as a reconstruction of the form of an Object. It follows that the Immediate Interpretant of a stretch of DNA or mRNA as a Composite Sign, i.e., its range of interpretability, amounts to the diversity of possibilities of reconstruction of the form of the Composite Immediate Object, the sequence of amino acids in a polypeptide. The Dynamical Interpretant of a stretch of DNA or mRNA corresponds to the effective reconstruction of a sequence of amino 72
Thus, when we talk about “range of interpretability” as what defines the Immediate Interpretant, two things should be emphasized: (a) This is a semiotic notion pointing to the complex contextual system of interpretation of the cell that co-determines the outcome of the particular process of protein synthesis (including alternative splicing), that is, that this process is both biochemical and semiotic. (b) It is an empirical issue to find out the precise interpretative mechanisms used by the cell in order to select a particular Dynamical Object (functional protein) out of many possible ones. These processes are not yet fully understood, and, therefore, it is not yet possible to describe the Immediate Interpretant on a more fine-grained level.
130
Genes, Information, and Semiosis
acids. In an alternatively spliced gene, such as the FN gene, this realization involves the instantiation of a specific splicing pattern in a given cell type, at a given developmental stage. Thus, one of the possibilities established in the range of interpretability of a stretch of DNA, in its Immediate Interpretant, is actualized. In a fibroblast, for instance, when a specific Immediate Object is synthesized, the fibroblast-specific FN isoform, this means that, from the range of possible sequences of amino acids that may be made out of the FN gene — its Immediate Interpretant —, a specific sequence is reconstructed — its Dynamical Interpretant. After it is actualized, an Immediate Object indicates a particular Dynamical Object — say, a specific FN isoform —, which the cell finds out through habits acquired in evolution and development. It is the Dynamical Object, then, that has an effect on the cell as a global interpreter. We can define, then, a Dynamical Interpretant of the Dynamical Object, a particular effect on a cell, among a range of possible effects — the Immediate Interpretant of the Dynamical Object. This Dynamical Interpretant is the actualization of one of the possible effects that a Composite Sign might have on the interpreter. Its range of interpretability is the Immediate Interpretant of the Composite Sign. See Table 1 for a summary of the correspondences between terms in the first take of our analysis. Table 1. Correspondence between biological and semiotic terms in the first take of our analysis. Semiotic term Sign Immediate Object
Dynamical Object - of a gene Composite Sign (formed by clusters of Simple Signs) Simple Signs
Biological term A string of DNA (a gene) A specific sequence of amino acids in a polypeptide or sequence of nucleotides in RNA A functional, folded, and chemically modified protein (e. g., a functional FN isoform) A stretch of DNA as a whole (or mature RNA) Codons
Chapter 6. A semiotic analysis of genes and genetic information
131
Semiotic term Interpretant Immediate Interpretant
Biological term A process of reconstruction of a specific form Range of interpretability of a Sign in DNA, i.e. the possibility of reconstruction of specific immediate objects
- of a codon as a Simple Sign
- the range of interpretability established by the rules of base pairing by which specific nucleotides in DNA determine specific nucleotides in mRNA or - the range of interpretability of threenucleotide sequences in mature mRNA by means of which nucleotide sequences determine the addition of specific amino acids to a growing polypeptide chain
- of a Composite Sign
One of the processes that enrich the range of interpretability Dynamical Interpretant - of a codon as a Simple Sign - of a stretch of DNA or mRNA as a Composite Sign The Composite Immediate Object of a protein-coding gene (= the Dynamical Object in its semiotically available form) a Simple Immediate Object
- the range of interpretability of that Sign in DNA, i.e., the fact that it can be interpreted so as to produce a range of potential Immediate Objects, potential sequences of amino acids that can be produced from that Sign in DNA RNA splicing
- the realization of one of the rules of base pairing or of the genetic code - the effective reconstruction of a sequence of amino acids The sequence of amino acids of a polypeptide
Each single amino acid
The analysis presented in this section faces the potential problem that it seems to treat the Sign as the primary constraining factor in semiosis, while this role is reserved for the Dynamical Object in
132
Genes, Information, and Semiosis
Peirce’s theory of Signs.73 After all, we are describing here how S (a sequence of nucleotides in DNA) determines O (a sequence of amino acids in a polypeptide) through I (a range of possibilities of reconstruction of sequences of amino acids).74 We can accommodate this description to a Peircean framework by examining the constraining action of the Object in evolutionary terms (see Figure 7). This perspective is related both to Darwin’s theory of natural selection and to the idea of causes not simply conceived mechanistically but as producers of general types of possible outcomes, or habits, i.e., final causes in the sense of Peirce (cf. Short 2002). Consider two different generations of a population, in times t1 and t2, and a set of tokens of a specific type of functional protein (Dynamical Object) in t1 that increases the likelihood of successful, adaptive experiences of organisms possessing it. Therefore, that set of proteins increases the likelihood that a gene (Sign) encoding them will be present in high frequencies in the next generation, in t2. This gene, in turn, will bring to the next generation the potency to produce tokens of that type of protein, as a Dynamical Object, by indicating it through its semiotically available form, its Immediate Object. Signs in DNA will carry to future generations the potentiality of reconstructing the form of that protein in generations to come. This means that that gene, as a Sign, exerts a determining influence on the range of possibilities of reconstructing sequences of amino acids in the next generation. If we follow this set of ideas, we will be able to see how, in evolutionary terms, O determines I through S, in conformity with Peirce’s account of semiosis. Nevertheless, the role of O as the primary constraining factor of semiosis depends, in the genetic information system, on the role of S, in a given generation, in determining O through I. We can say, in short, that the fact that S determines O through I in a given population in t2 is itself determined by the fact that O determined I by 73
The irreducibility of the triadic relation S-O-I is a logical property. Therefore, while it makes no sense to sort out a primary constraining factor in such a logical relation, dynamically it makes sense to sort out the Dynamical Object as the primary constraining factor of semiosis (for a detailed discussion about this issue, see Short 1998: 31). 74 In this picture, it is important to take in due account that we are not claiming that DNA causes or brings about the protein as an Object, since DNA is a set of data (or, as we prefer, signs) rather than a program, a source of materials rather than a master agent in the cell. It is the DNA processing system that produces the proteins. We are not claiming, therefore, that the Sign causes the Object.
Chapter 6. A semiotic analysis of genes and genetic information
133
increasing the likelihood of S being present in a high frequency in t2, by means of its involvement in successful experiences in t1.
Figure 7. The Dynamical Object (functional protein) as the primary constraining factor of semiosis in the genetic information system. S, Sign; DO, Dynamical Object, IO, Immediate Object; II, Immediate Interpretant; t, generation time.
As Maynard Smith (2000a: 177) puts it, the sequence of a gene is determined, by past natural selection, because of the effects, or rather, the types of effects, it produces. This is a case where types of outcomes (generals) are involved in explaining why there are those types of outcomes. The patterns of outcomes are explanatory by being ‘selected for’, and such an explanation is, as Short (2002) has described, nontrivially teleological. The relationship between Signs in DNA and the sequence of amino acids of a protein (the Composite Immediate Object) is established by a complex mechanism of interpretation, involving transcription, RNA processing and translation. Thus, to interpret a string of DNA, more than one interpretative system is required, including, for instance, RNA polymerases, involved in the transcription of DNA into RNA, and ribosomes, involved in the translation of mRNA into proteins. These interpretative systems are parts or subsystems of a cell as a global interpreter, and their actions are subordinated to the latter. The idea that the cell can be seen as a global interpreter to which a series of interpretative subsystems in the genetic information system are subordinated is dramatically reinforced by recent analyses of the functional organization of proteomes. For instance, Gavin et al. (2002, 2006) showed that the vast majority of protein complexes in yeast are
134
Genes, Information, and Semiosis
associated with one another, directly or indirectly, through common proteins. As a researcher told Sampedro (2004: 61), it is as if ‘the whole cell was a single machine.’ More than half of the protein complexes analyzed by Gavin et al. are involved in the genetic information system: transcription/DNA maintenance/chromatin structure (24%); RNA metabolism (12%); protein synthesis/turnover (14%); signaling (9%); and protein/RNA transport (5%). Even more interestingly, the multi-component cellular systems involved in transcription, RNA processing, and RNA transport do not form a simple linear assembly line, but a complex and extensively coupled network in which signals circulate in a non-linear manner, involving several feedback loops (Maniatis, Reed 2002; Kornblihtt et al. 2004). It is this network structure which makes it possible the coordination of the interpretative subsystems in the genetic information system by the cell. It is clear, then, that we cannot easily move from claims at the cell level to claims at the molecular level while pondering about which system is interpreting genes as Signs. We think that these recent studies clearly show that, when a gene is interpreted, the interpretation process is indeed taking place at the cellular level, albeit multi-component molecular subsystems are necessary to this endeavor. The idea that ultimately the whole cell participates in the network necessary for the interpretation that is demanded for the effect of a gene product to take place (cf. Emmeche, Hoffmeyer 1991; Pardini, Guimarães 1992) is further supported by the role of an impressive array of signaling pathways regulating the interpretation of Signs in DNA. As Fogle (2000: 19) sums up, “DNA action and function become meaningful in the context of a cellular system. Coding information in the DNA is necessary but insufficient for the operation of living systems.” A Peircean approach to the gene concept entails that genetic structures should not be seen in isolation from the larger system by which they are interpreted. From this perspective, the meaning of a gene to its interpreter, the cell, or, to put it differently, the biological meaningfulness of a gene, is found not only in DNA sequences in a chromosome. After all, there is more to genetic information than just a sequence of nucleotides in DNA. We will have to include the effect of the gene-as-a-Sign on the cell or organism, and, in fact, the very role of cellular subsystems as interpreters of strings of DNA, in such a way that they relate Signs to specific Dynamical Objects, proteins which
Chapter 6. A semiotic analysis of genes and genetic information
135
play a function inside the cellular system and have an effect on it or on the organism of which the cell is a part. The traditional identification of genetic information with sequential information in DNA molecules made it impossible to understand it as a triadic-dependent, semiotic process, as we propose here. In other words, in the classical molecular gene concept, information was often considered to be simply reduced to its vehicle, DNA, isolated from all the other elements in what we analyze as a triadic process that comprises the action of a gene as a Sign. We propose here that we should regard information as precisely this action of a gene as a Sign, understanding it as a process including more elements than just Signs in DNA. In our view, this first-take semiotic analysis of the genetic information system leads to the following conclusions (Queiroz et al. 2005; El-Hani et al. 2006):75
75
It is important to stress that the semiotic analysis we performed does not entail that analogies between genetic sequences and human languages can be taken too far, as we see, for instance, in the following passage: “Consider all of the structural information required to build a polypeptide chain and all of the regulatory information required to deploy that polypeptide in the correct sets of cells at the proper developmental times and in the requisite quantities. If every set of such information were analogous to one sentence in the instruction manual that we call the genome, a reasonable current assessment is that we have a partial but still quite incomplete knowledge of how to identify and read certain nouns (the structures of the nascent polypeptides and protein-coding exons of mRNAs). Our ability to identify the verbs and adjectives and other components of these genomic sentences (for example, the regulatory elements that drive expression patterns or structural elements within chromosomes) is vanishingly low. Further, we do not understand the grammar at all — how to read a sentence, how to weave the different sentences together to form sensible paragraphs describing how to build multicomponent proteins and other complexes, how to elaborate physiological or developmental pathways, and so on” (Gelbart 1998: 659). These statements are merely metaphorical and they remain so in face of the analysis developed here. It is not in this sense that we believe that a semiotic analysis can help turning information talk more than metaphorical, but in the formal, abstract sense in which Peirce’s theory of Signs allow us to conceive the relationship between Signs, Objects, and Interpretants in the genetic system, as argued above. As Guimarães and Moreira (2000) argue, significant segments in DNA are not organized as substantives, predicates, verbs, adverbs, etc., with syntax similar to those of human languages.
136
Genes, Information, and Semiosis
Genes should be regarded as Signs in DNA, which can only have any effect on a cell through a triadic-dependent process (semiosis); this process is genetic information and involves more than just genes as Signs in DNA but also Objects and Interpretants; genetic information is the process by means of which a form in a Dynamical Object (a functional protein) is communicated to an Interpretant (the reconstruction of a specific sequence of amino acids in a cell) through Signs in DNA.
In the next chapters, we will turn to a more detailed analysis of some processes in the genetic information system. The conclusions presented above will be both substantiated and extended in significant ways by this more fine-grained analysis.
Chapter 7 Emergence of semiosis: a general model
1. A multi-level approach to the emergence of semiosis in semiotic systems In order to refine and extend the semiotic analysis developed in the previous chapter, we will elaborate a case study of some processes involved in the genetic information system. As a first step in this case study, we will propose a model for treating semiosis in cellular (and other kinds of) systems as involving relationships at several levels. In this model, we address semiotic processes at three levels at a time, on the grounds of Salthe’s (1985) ‘basic triadic system’, clearly influenced by Peirce (see also Queiroz, El-Hani 2006a, 2006b). The basic triadic system plays a fundamental role in Salthe’s ‘hierarchical structuralism’, conceived by him as a coherent and heuristically powerful way of representing natural entities. This role follows from the prospect of discovering by means of this system general rules and principles of constraint within which the laws of nature must operate. According to the basic triadic system, to describe the fundamental interactions of a given entity or process in a hierarchy, we need (i) to consider it at the level where we observe it (‘focal level’); (ii) to investigate it in terms of its relations with the parts described at a lower level (usually, but not necessarily always, the next lower level); and (iii) to take into account entities or processes at a higher level (also usually but not always the next higher level), in which the entities or processes observed at the focal level are embedded. Both the lower and the higher levels have constraining influences over the dynamics of the entities and/or processes at the focal level. These constraints allow us to explain the emergence of entities or processes (e. g., semiosis) at the focal level. In a manner which is consistent with Peircean pragmatism, the choice of the focal, lower, and higher levels depends on the research goals. Therefore, it results from a decision made by a researcher on the grounds of a theoretical framework and methodological approach. The theoretical and methodological bases chosen by a researcher, in turn, are partly (and, often, strongly) influenced by the epistemic practices accepted as scientifically adequate and, typically, also standardized by the scientific community. Therefore, a researcher can
140
Genes, Information, and Semiosis
choose as a focal level in her investigation a level in which another researcher, guided by another purpose, locates the boundary conditions in the triadic system she is studying. We will see below that a higher level constraining focal-level semiotic processes can include itself semiotic processes. The latter would turn into focal-level processes for a research aiming specifically at studying them. At the lower level, the constraining conditions amount to the ‘possibilities’ or ‘initiating conditions’ for the emergent process, while constraints at the higher level are related to the role of a selective environment played by the entities at this level, establishing the boundary conditions that coordinate or regulate the dynamics at the focal level.76 In this model, an emergent process at the focal level is explained as the product of an interaction between processes taking place at lower and higher levels. The phenomena observed at the focal level should be ‘… among the possibilities engendered by permutations of possible initiating conditions established at the […] lower level’ (Salthe 1985: 101). Nevertheless, processes at the focal level are embedded in a higher-level environment that plays a role as important as that of the lower level and its initiating conditions. Through the temporal evolution of the systems at the focal level, this environment or context selects among the states potentially engendered by the components at the lower level those that will be effectively actualized. As Salthe (1985: 101) puts it, “what actually will emerge will be guided by combinations of boundary conditions imposed by the […] higher level”. Figure 8 shows a scheme of the determinative relationships in Salthe’s basic triadic system.77 76
The regulation of a focal-level process by higher-level boundary conditions is interpreted here as a kind of selective process. Suppose that the causal relation between a given element of a system, A, and another element of the same system, B, is regulated. This is understood, in this framework, as the selection of B as the most probable effect of A, among other possible effects, by boundary conditions established by a level higher to the level where the causal relation at stake is taking place. This is related to ideas found in Polanyi (1968), who introduced the term ‘boundary conditions’ in the biological sciences, and Campbell (1974), who introduced the expression ‘downward causation’, commonly employed in discussions about emergence (see below). 77 We systematically use the concept of emergence in this section — as Salthe himself did – in connection with the basic triadic system and, furthermore, with our treatment of semiotic processes. Nevertheless, in order to use this concept and its derivatives in a consistent way, one needs to address the conditions which
Chapter 7. Emergence of semiosis: a general model
141
Figure 8. A scheme of the determinative relationships in Salthe’s basic triadic system. In a perfectly nested hierarchy, the focal level is not only constrained by boundary conditions established by the higher level, but also establishes the potentialities for constituting the latter. In turn, when the focal level is constituted from potentialities established by the lower level, a selection process also takes place, since among those potentialities some will be selected in order to constitute a given focal-level process.
For the sake of the argument, let us begin by taking as the ‘focal level’ that level in which a selected semiotic process is observed. Semiotic processes at the focal level are described here as chains of triads. We can treat, then, the interaction between semiotic processes at the focal level, potential determinative relations between elements at a lower level (‘micro-semiotic level’), and semiotic processes at a higher level (‘macro-semiotic level’). In the latter, networks of chains of triads which embed the semiotic process at the focal level are described. The micro-semiotic level concerns the relations of determination that may take place within each triad S-O-I. The relations of determination provide the way the elements in a triad are arranged in semiosis. According to Peirce, the Interpretant is determined by the Object through the mediation of the Sign (I is determined by O through S) (Peirce MS 318: 81). This is a result from two determinative relations: the determination of the Sign by the Object relatively to the Interpretant (O determines S relatively to I), and the determination of the Interpretant by the Sign relatively to the Object (S determines I relatively to O) (Tienne 1992). should be fulfilled for a process or property to be characterized as ‘emergent’. We will discuss these conditions later in this chapter.
142
Genes, Information, and Semiosis
In the micro-semiotic level, we consider that, given the relative positions of S, O, and I, a triad ti = (Si, Oi, Ii) can only be defined as such in the context of a chain of triads T = {..., ti-1, ti, ti+1,...} (see Gomes et al. 2003; Queiroz, El-Hani 2006a,b). A Sign in action, i.e., semiosis, entails the instantiation of chains of triads. As Savan (1986: 134) argues, an Interpretant is both the third term of a given triadic relation and the first term (Sign) of a subsequent triadic relation. This is the reason why semiosis cannot be defined as an isolated triad; it necessarily involves chains of triads (see Merrell 1995) (see Figure 9). Given the framework of Salthe’s hierarchical structuralism, we should analyze semiosis by considering three levels at a time: Each Sign action is modeled as a chain of triads located at a focal level, and, correspondingly, we will talk about focal-level semiotic processes. Micro-level semiotic processes will involve the relations of determination within each triad. Macro-level semiotic processes will involve networks of chains of triads, in which each individual chain is embedded. Focal-level semiosis will emerge as a process through the interaction between micro- and macro-semiotic processes, i.e., between the relations of determination within each triad and the embedment of each individual chain in a whole network of Sign processes.
Figure 9. Scheme showing that a triad can only be defined within a chain of triads. The grid at the bottom part of the figure shows that Oi-1, Oi, and Oi+1 are Immediate Objects of the same Dynamical Object.
Chapter 7. Emergence of semiosis: a general model
143
Following Salthe’s explanation of constraints, micro-semiosis establishes the initiating conditions for focal-level semiotic processes. Each chain of triads always indicates the same Dynamical Object through a series of Immediate Objects, as represented in each triad (see Figure 9). The potentialities of indicating a Dynamical Object are constrained by the relations of determination within each triad. That is, the way O determines S relatively to I, and S determines I relatively to O, and then how I is determined by O through S leads to a number of potential ways in which a Dynamical Object may be indicated in focal-level semiosis, i.e., to a set of potential triadic relations between Immediate Objects, Signs, and Interpretants. We need to consider, then, the distinction between potentiality and actuality in the context of our analysis. For this purpose, we introduce the definitions of potential Signs, Objects, and Interpretants. A ‘potential Sign’ is something that may be a Sign of an Object to an Interpretant, i.e., it may stand for that Object to an Interpretant. A ‘potential Object’, in turn, is something that may be the Object of a Sign to an Interpretant. And, finally, a ‘potential Interpretant’ is something that may be the Interpretant of a Sign, i.e., it may be an effect of that Sign. The micro-semiotic level is the domain of potential Signs, Objects, and Interpretants. Then, we can consider a whole set W of possible determinative relations between these three elements, which can generate, in turn, a set of possible triads. These triads cannot be fixed, however, by the micro-semiotic level, since the latter establishes only the initiating conditions for chains of triads at the focal level. To fix a chain of triads, and, consequently, the individual triads defined within that chain, boundary conditions established by the macro-semiotic level should also play their selective role. That is, networks of chains of triads constitute a semiotic environment or context which plays a fundamental role in the actualization of potential chains of triads. Chains of triads are actualized at the focal level by a selection of those triads which will be effectively actualized amongst those potentially engendered at the micro-semiotic level. After all, a triad ti = (Si, Oi, Ii) cannot be defined atomistically, in isolation, but only when embedded within higher-level structures and/or processes, including both chains
144
Genes, Information, and Semiosis
of triads T = {..., ti-1, ti, ti+1,...} and networks of chains of triads ST = {T1, T2, T3,..., Tn}.78 It is in this sense that the emergence of semiotic processes at the focal level, in which chains of triads are actualized, is explained in this model as resulting from an interaction between the potentialities established by the micro-semiotic level and the selective, regulatory influence of the macro-semiotic level. The general ideas involved in this model of the emergence of semiosis are shown in Figure 10.
78 Even though we will not pursue this issue here, we should emphasize that we think it is obviously possible to construct a clear correspondence between the hierarchical structure proposed by Salthe and Peirce’s categories. The micro-semiotic level — at which processes relating S, O, and I are initiated — gives Sign processes an inevitable character of indeterminacy. It is straightforward, then, to associate the micro-semiotic level with firstness. Salthe himself stresses that this level exhibits a fundamentally stochastic behavior. At the focal level, specific, particular processes are spatiotemporally instantiated, as tokens, which are cases of secondness. The macro-semiotic level, in turn, gives Sign processes their generality and temporality, making them historical and context-dependent. We can say, thus, that the macro-semiotic level shows the nature of thirdness. Potentialities at the micro-semiotic level are not the same as mere possibilities. For the sake of our arguments, consider Peirce’s treatment of Quality as a ‘mere abstract potentiality’ (CP 1.422). Quality has the nature of firstness, being essentially indeterminate and vague. But we can also talk about a generality of Quality. In this case, we are beyond the realm of pure firstness, as generality refers to some law-like tendency, and thus to the nature of thirdness. Peirce works, in this case, with a merging of firstness and thirdness. It is in this latter sense that we understand potentialities at the micro-semiotic level here, as a particular set of potential Signs, Objects, and Interpretants which have been established due to the fact that the micro-semiotic level is embedded in a hierarchical system which includes levels showing the nature of secondness and thirdness (focal and macro-semiotic levels, respectively). These potentialities show, thus, the nature of a generality, being closer to a merging of firstness and thirdness than to pure firstness. Such a treatment seems to be compatible with Peirce’s categoreal scheme, since, as Potter (1997: 94) stresses, ‘the categoreal structure which Peirce uses is […] highly subtle and complex, admitting of various combinations.’
Chapter 7. Emergence of semiosis: a general model
145
Figure 10. A model of semiosis in three levels. The upward arrow shows the constitutive relation from individual triads to chains of triads, corresponding to Salthe’s initiating conditions. The downward arrow shows selective relations from networks of chains of triads to chains of triads, corresponding to Salthe’s boundary conditions.
2. Semiosis as an emergent process In the previous section, the concept of emergence and its derivatives were used in connection with Salthe’s basic triadic system and our analysis of semiotic processes. In order to apply these concepts consistently, we need to bring into play a technical treatment of emergence and, also, discuss if and in what sense semiosis can be regarded as an ‘emergent’ process in semiotic systems. In the end of the 1980s, emergentism seemed to be an entirely forgotten philosophical position. For instance, in 1990, Kim wrote that “the emergence debate [...] has by and large been forgotten, and appears to have had negligible effects on the current debates in metaphysics, philosophy of mind, and philosophy of science [...]” (in Kim 1993: 134). Nevertheless, the fortunes of emergentism have definitely changed in the last decade, as we can see in the way Kim perceived its situation in the philosophical landscape only some years after declaring its oblivion. After recognizing that emergentism was coming back into the philosophical scene, Kim (1997: 271) describes a picture in which we can “see an increasing, and unapologetic, use of
146
Genes, Information, and Semiosis
expressions like ‘emergent property’, ‘emergent phenomenon’, and ‘emergent law’, substantially in the sense intended by classic emergentists, not only in philosophical writings but in primary scientific literature as well” (Kim 1999: 4). And this situation goes on to the present day. We can really say now that the debate about emergence has re-emerged (Kim 1998, 1999; Stephan 1999; Cunningham 2001; Pihlström 2002; El-Hani 2002a). A great number of works dealing with emergence have been published in the philosophical and scientific literature in the last 15 years.79 The concept of emergence has been increasingly used in so diverse fields as artificial life, cognitive sciences, evolutionary biology, theories of self-organization, philosophy of mind, dynamical systems theory, synergetics, etc. The role played by the concept in these fields has been directly responsible for revitalizing emergentism as a philosophical trend. This seems quite natural, since many of these fields can be gathered under a general description as the ‘sciences of complexity’, concerned with ‘the complex emerging properties of life and mind’ (Emmeche 1997; 1998). In recent years, the very term ‘emergence’ (and its derivatives) has been widely used in research fields largely based on computer simulations, such as artificial life, cognitive robotics, and synthetic ethology (Cariani 1989; Emmeche 1994, 1997, 1998; Ronald et al. 1999; MacLennan 2001; Bedau 2002; Cangelosi, Turner 2002). In these fields, the concept of ‘emergence’ has indeed become so popular that they are often described as dealing with ‘emergent computation’. In the case of Alife, Langton (1989: 2) even states that the key concept in this field is that of ‘emergent behavior’. Surprisingly, little discussion regarding the precise meaning of the terms ‘emergence’, ‘emergent’, and so on, is found in them, and, furthermore, in ‘emergent’ computation, they are often used in vague and imprecise ways, 79
E. g., Klee (1984); Savigny (1985); Blitz (1992); Beckermann et al. (1992); Stephan (1997, 1998, 1999); Kim (1997, 1999); O’Connor (1994); Baas (1996); Newman (1996); Baas, Emmeche (1997); Humphreys (1996, 1997a, 1997b); Emmeche et al. (1997); Emmeche (1997); Bedau (1997, 2002); Azzone (1998); Schröder (1998); El-Hani, Pereira (1999); Pihlström (1999, 2002); El-Hani, Emmeche (2000); Andersen et al. (2000); El-Hani, Videira (2001); El-Hani, Pihlström (2002a, 2002b; 2005); El-Hani (2002a); Symons (2002); Gillett (2002). Emergence has also been the topic of recent special issues of a series of journals, such as Philosophical Studies (1999), International Journal of Systems Science (2000), Grazer Philosophische Studien (2002), Principia (2002), and Synthese (2006).
Chapter 7. Emergence of semiosis: a general model
147
without a rigorous debate concerning how they might be explained or defined (for critical commentaries, see Cariani 1989; Bedau 2002; ElHani 2002a). In fact, the price of being careless about the use of this concept is already surfacing in the current debates about emergence, in the form of a certain perplexity about what is really meant by ‘emergence’, ‘emergent properties’, and so on. Given all the current debates about emergence and the important role this concept has come to play in many research fields, we should try to keep its meaning clear, inasmuch as it has carried for a long time a burdensome load of confusion about its metaphysical and epistemological aspects. There is still great confusion about what is really meant by ‘emergence’, ‘emergent properties’ etc., and, consequently, we are entitled here to ascribe a precise meaning to these concepts as we apply them to the domain of semiotic phenomena. In a broader sense, we also believe that a systematic approach to this topic is crucial in the domain of semiotic investigations in general, particularly as regards the debates in the context of Peirce’s metaphysics and evolutionary cosmology. Notably, we do not find in these debates (e. g., Parker 1998; Nöth 1994; Kruse 1994) technical or detailed discussions of the possible relationships between the concepts of semiosis and emergence. It is important to be clear from the start about our goals in this section. We do not intend to answer here when or how semiosis emerged in the universe. Our aim is to discuss conditions which should be fulfilled for semiosis to be characterized as an emergent process. For this purpose, we will presuppose, in our arguments, the existence of semiotic systems, in which semiosis is instantiated, but this doesn’t mean we are taking any stance regarding the possibility that semiosis may have in some sense preceded such systems. We will begin our treatment of emergence by summarizing a systematic analysis of the variety of emergence theories and concepts developed by Stephan (1998, 1999). This will lead us to pose fundamental questions that have to be answered in order to ascribe a precise meaning to the term ‘emergence’ in the context of semiotics. Finally, we will propose some tentative answers to the questions raised along the presentation of Stephan’s analysis.
148
Genes, Information, and Semiosis
2.1. Central characteristics of emergentism: what questions about semiosis do they raise? In this book, we intend to apply the concept of emergence to the domain of semiotic phenomena in a precise manner, with the purpose of stimulating a more thoroughgoing dialogue between the philosophical traditions of semiotics and emergentism. For the sake of our arguments, we will employ Stephan’s systematic analysis of emergence theories and concepts as a basis for posing fundamental questions that have to be answered in order to characterize semiosis as an ‘emergent process’. The term ‘emergence’ is often employed in an intuitive and ordinary way, referring to the idea of “creation of new properties”. This idea comes back to one of the original sources of the emergentist thinking, the works of the British psychologist Conwy Lloyd Morgan. As shown by Emmeche et al. (1997), a discussion of the key concepts in this idea, ‘novelty’, ‘property’, and ‘creation’, can result in an understanding of some of the main issues in emergentism. Nevertheless, this idea is not enough for grasping the concept of emergence, mainly because it is focused on characteristic claims of one type of emergentism, namely, ‘diachronic emergentism’ (see below). In a technical sense, ‘emergent’ properties can be understood as a certain class of higher-level properties related in a certain way to the microstructure of a class of systems (adapted from Stephan 1998: 639). The reason why such a broad definition, with open clauses, seems at first more adequate than a definition with more content and precision has to do with the fact that the concept of emergence and its derivatives are employed in the most diverse fields, and, consequently, a more detailed definition is likely to apply to some fields but not to others. It is part of the task of an emergence theory — when applied to a particular research field — to fill in the open clauses in this definition (shown in italics). An emergence theory should, among other things, provide an account of which properties (of a given class of systems) should be regarded as ‘emergent’, and offer an explanation of the relationship between these properties and the microstructure of the systems in which they are instantiated. Moreover, it should establish which systems exhibit a certain class of emergent properties. If we extend the definition above to encompass processes, a first question to
Chapter 7. Emergence of semiosis: a general model
149
be answered in order to characterize semiosis as an emergent process concerns the demarcation of the class of systems which show semiosis. We can frame it as follows: (1) what is a semiotic system? There is no unified emergence theory. Rather, emergence theories come in various shapes and flavors. Nevertheless, it is possible to recognize in the diversity of emergence theories a series of central characteristics (Stephan 1999: ch. 3; cf. also Stephan 1998). First, emergentists should, in a scientific spirit, be committed to some form of naturalism, claiming that only natural factors play a causal role in the evolution of the universe. Even though naturalism and materialism (or, for that matter, physicalism) philosophically do not coincide80, it is the case that, in the current scientific picture, a naturalisticallyminded emergentist should stick to the idea that all entities consist of physical parts. This thesis can be labeled ‘physical monism’: there are, and will always be, only physically constituted entities in the universe, and any emergent property or process is instantiated by systems that are physically constituted.81 Therefore, we can pose the following question: (2) are semiotic systems physically constituted? A second characteristic mark of emergentism is the notion of novelty: new systems, structures, processes, entities, properties, and dispositions are formed in the course of evolution. This idea entails the following question: (3) do semiotic systems constitute a new class of systems, instantiating new structures, processes, properties, dispositions, etc.? Emergence theories require, thirdly, a distinction between systemic and non-systemic properties. A property is systemic if and only if it is found at the level of the system as a whole, but not at the level of its parts. Conversely, a non-systemic property is also observed at the parts of the system. If we similarly propose a distinction between systemic and non-systemic processes, the next question can be raised: (4) can semiosis be described as a systemic process?
80
About differences between naturalism and physicalism, see, e. g., Caro, MacArthur (2004). 81 This thesis is what Ernst Mayr, a co-founder of the modern theory of evolution and himself an organicist/emergentist, called in a biological context “constitutive reductionism” (Mayr 1982: 60), i.e., the uncontroversial idea that the material composition of organisms is exactly the same as that found in the inorganic world.
150
Genes, Information, and Semiosis
A fourth characteristic of emergence theories is the assumption of a hierarchy of levels of existence. Thus, it is also necessary to answer the following question to convincingly characterize semiosis as an emergent process: (5) how should we describe levels in semiotic systems and, moreover, how do these levels relate to the emergence of semiosis? A fifth characteristic is the thesis of synchronic determination, a corollary of physical monism: a system’s properties and behavioral dispositions depend on its microstructure, i.e., on its parts’ properties and arrangement; there can be no difference in systemic properties without there being some difference in the properties of the system’s parts and/or in their arrangement. The next question to be addressed, then, is the following: (6) in what sense can we say (and explain) that semiosis, as an emergent process in semiotic systems, is synchronically determined by the properties and arrangement of their parts? Sixthly, although some emergentists (e.g., Popper in Popper, Eccles, [1977]1986) have subscribed to indeterminism, one of the characteristics of emergentism (at least in the classical British tradition) is a belief in diachronic determination: the coming into existence of new structures would be a deterministic process governed by natural laws (Stephan 1999: 31). This is certainly one feature of classical emergence theories which is incompatible with Peirce’s theoretical framework, as he rejected the belief in a deterministic universe (CP 6.201). But this does not preclude the treatment of emergence in connection to a Peircean account of semiosis, as there are also emergence theories committed to indeterminism. It is not necessary at all to be imprisoned in the old British tradition of emergentist thought. Seventhly, emergentists are committed to the notion of the irreducibility of a systemic property designated as ‘emergent’. An eighth important notion is that of unpredictability (in principle). We should, then, pose two more questions: (7) in what sense can we say that semiosis, as observed in semiotic systems, is irreducible? (8) in what sense can we claim that the instantiation of semiosis in semiotic systems is unpredictable in principle? Finally, the ninth characteristic of emergentism is the idea of downward causation: novel structures or new kinds of states of ‘relatedness’ of preexistent objects manifest downward causal efficacy, determining the behavior of a system’s parts. Given this idea, yet another question should be raised: (9) is some sort of downward
Chapter 7. Emergence of semiosis: a general model
151
causation involved in semiosis? We will discuss these latter notions in a fine-grained manner in the next section.
2.2. Varieties of emergentism and questions about semiosis Several different emergence theories have been proposed throughout the 20th century. The characteristic marks discussed above allow one to define several varieties of emergentism, significantly differing from one another in strength (see Stephan 1998; 1999: Ch. 4). For the sake of our arguments, we will consider just three basic varieties of emergentism — weak; synchronic; and diachronic.82 Weak emergentism assumes (1) physical monism, (2) a distinction between systemic and non-systemic properties, and (3) synchronic determination. This comprises the minimal conditions for a physicalist emergentist philosophy. Thus, weak emergentism is the common basis for all stronger physicalist emergence theories. However, this view in itself is weak enough to be compatible with reductive physicalism (Stephan 1998: 642; 1999: 67). Consequently, weak emergentism faces a fundamental problem as regards the basic motivations underlying the efforts of most emergence theorists, who typically take emergentism to be by definition an anti-reductionist stance. It is just natural, then, that many emergentist philosophers and scientists are engaged in attempts to build stronger accounts of emergence, which are arguably incompatible with reductionism. Nevertheless, it is in this move that emergentist philosophies find the most difficult problems. Both synchronic and diachronic emergentism comprise strong emergence theories. In this work, we intend to characterize semiosis as an emergent process in a strong sense. Therefore, we have to analyze in more detail the concepts of “irreducibility” and “unpredictability”, assumed in synchronic and/or diachronic emergentism. Synchronic and diachronic emergentism are closely related, being often interwoven in single emergence theories, but, for the sake of clarity, it is important to distinguish between them. Synchronic emergentism is primarily interested in the relationship between a system’s 82
Stephan begins his systematic analysis of varieties of emergentism by discussing these three basic forms, but later expands his typology to include six different emergentist positions. It is outside the scope of this book to deal with all six positions. For more details, see the original works.
152
Genes, Information, and Semiosis
properties and its microstructure. The central notion in synchronic emergentism is that of irreducibility. Diachronic emergentism, by its turn, is mainly interested in how emergent properties come to be instantiated in evolution, focusing its arguments on the notion of unpredictability. No strong emergence theory can be properly formulated without coming to grips with the problems of irreducibility and/or unpredictability.
Modes of irreducibility The thesis of the irreducibility of emergent properties makes synchronic emergentism incompatible with reductive physicalism. But what does it precisely mean to say that a property is ‘irreducible’? Many discussions about emergence and reduction appeal to a rather generic reference to the term ‘reduction’, without considering that this term has a variety of meanings in the philosophical (and also scientific) literature (El-Hani, Emmeche 2000; El-Hani 2004; El-Hani, Queiroz 2005b). Therefore, there should be also different senses in which an emergent property can be treated as ‘irreducible’. Stephan (1998: 642–643; 1999: 68) distinguishes between two kinds of irreducibility. The first notion is based on the unanalyzability of systemic properties (adapted from Stephan 1998: 643): (I1) [Irreducibility as unanalyzability] Systemic properties which cannot be analyzed in terms of the behavior of a system’s parts are necessarily irreducible.
This notion plays an important role in the debates about qualia and is related to a first condition for reducibility, namely, that a property P will be reducible if it follows from the behavior of the system’s parts that the system exhibits P. Conversely, a systemic property P of a system S will be irreducible if it does not follow, even in principle, from the behavior of the system’s parts that S has property P. A second notion of irreducibility is based on the non-deducibility of the behavior of the system’s parts (adapted from Stephan 1998: 644): (I2) [Irreducibility of the behavior of the system’s parts] A systemic property will be irreducible if it is synchronically determined by the specific behavior the parts show within a system of a given kind, and this behavior, in turn, does not follow from the parts’ behavior in isolation or in other (simpler) kinds of system.
Chapter 7. Emergence of semiosis: a general model
153
This mode of irreducibility is related to the notion of downward causation, since it is plausible to assume that the influence of the system where a given emergent property P is observed on the behavior of its parts is the reason why we are not able to deduce the latter from the behaviors the very same parts show in isolation or as parts of simpler systems. A second condition for reducibility is violated in this case, which entails that a systemic property P of a system S will be irreducible if it is realized by parts of S whose behaviors do not follow, even in principle, from their behavior in systems simpler than S. Recently, Boogerd and colleagues (2005) explained the notions of irreducibility as unanalyzability and as non-deducibility in terms of two conditions for emergence they call “vertical” and “horizontal”. They take Broad’s works (1919, 1925) as a starting point to distinguish between these two independent conditions, which Broad himself did not explicitly differentiate (Figure 11). A systemic property PR of a system R(A,B,C) is emergent if either of these conditions is fulfilled. The vertical condition captures the situation in which a systemic property PR is emergent because it is not explainable, even in principle, with reference to the properties of the parts, their relationships within the entire system R(A,B,C), the relevant laws of nature, and the required composition principles. The horizontal condition grasps the situation in which a systemic property PR is emergent because the properties of the parts within the system R(A,B,C) cannot be deduced from their properties in isolation or in other wholes, even in principle. Since these two conditions are independent, there are two different possibilities for the occurrence of emergent properties: (i) a systemic property P of a system S is emergent if it does not follow, even in principle, from the properties of the parts within S that S has property P; and (ii) a systemic property P of a system S is emergent, if it does not follow, even in principle, from the properties of the parts in systems different from S how they will behave in S, realizing P. The vertical condition for emergence expresses in a different way the idea of unanalyzability. Even if we know (i) what properties and relations A, B, and C show within the system R(A,B,C), (ii) the relevant laws of nature, and (iii) all necessary composition principles, yet we will not be able to deduce that the system has property P. This
154
Genes, Information, and Semiosis
is a case in which the condition of analyzability is violated, since it does not follow, even in principle, from the behavior of the parts A, B, and C in system R(A,B,C) that the system has P.
Figure 11. Vertical and horizontal conditions for emergence. A, B, and C are the parts making up the system R(A,B,C), which shows PR, a systemic property. S1(A,B), S2(A,C), and S3(B,C) are simpler systems including these parts. T1(A,B,D) is a system with the same number of parts, and T2(A,C,D,F) is a system with more parts than R(A,B,C). The diagonal arrow represents Broad’s idea of emergence. The horizontal and vertical arrows capture the two conditions implicit in Broad that Boogerd et al. make explicit. (From Boogerd et al. 2005).
The horizontal condition for emergence expresses in a different way the idea of irreducibility based on the non-deducibility of the behavior of the system’s parts. In this case, if we know the structure of the system R(A,B,C), we will be able to explain and predict the behavior of the parts within it, and, also, the instantiation of property P. Boogerd et al. (2005) discuss the resources available for deducing the behavior of the parts within R(A,B,C) from their properties and behaviors in other kinds of systems, in order to establish what would be the proper basis for this inference. We may deduce the behavior of the parts in R(A,B,C) from their properties and behaviors in systems of greater, equal or less complexity (see Figure 11).83 The possible bases 83 It is important to stress that Boogerd et al. (2005) are obviously aware that complexity does not depend only on the number of components of a system, but
Chapter 7. Emergence of semiosis: a general model
155
for deduction of the parts’ behavior in R(A,B,C) include: (i) more complex systems, such as T2(A,C,D,F); (ii) systems with the same degree of complexity, such as T1(A,B,D); (iii) simpler systems, such as S1(A,B), S2(A,C), and S3(B,C); and (iv) the parts A, B, and C in isolation. They convincingly argue that only (iii) is an interesting basis for deduction, since (iv) trivializes emergence (i.e., in this case, each and every property of a system would seem to be ‘emergent’), and (i) and (ii) trivialize non-emergence (i.e., in this case, each and every property of a system would seem to be ‘non-emergent’). They conclude, thus, that the key case for understanding the horizontal condition for emergence is (iii), in which we attempt to deduce the behavior of R(A,B,C) or its parts on the basis of less complex systems. A more fine-grained analysis of the irreducibility concept naturally leads to a reframing of the seventh question raised above: (7) Which interpretation of irreducibility is more adequate to understand Peirce’s claims about the irreducibility of semiosis? Furthermore, the explanation of irreducibility as non-deducibility makes it evident that question 9, “is some sort of downward causation involved in semiosis?”, should be posed in connection with this particular interpretation. This raises a number of difficult questions. Symptomatically, the problem of downward causation (DC) is one of the most debated in the contemporary literature on emergence (see, e. g., Schröder 1998; Stephan 1999; Andersen et al. 2000; El-Hani, Emmeche 2000; El-Hani 2002; Hulswit 2005). We will not pursue this debate here in all its details. Rather, we will discuss some central ideas and controversies about DC, in order to subsequently consider them with regard to semiotic phenomena.
Downward determination Emmeche and colleagues (2000) presented a systematic analysis of different notions of DC, identifying three versions, each making use of a particular way of interpreting the causal mode (or modes) involved in the influence of a whole over its parts: strong, medium, and weak DC. Strong DC interprets the causal influence of a whole over its parts also on its structure and the nature of the interactions among the parts. They indicate differences in complexity through number of parts only for the sake of the argument.
156
Genes, Information, and Semiosis
as a case of ordinary, efficient causation. Nevertheless, to claim that a higher level exerts an efficient causal influence over a lower one, we need to postulate that there is a sharp distinction between them, regarding them ultimately as being constituted by different kinds of substances (Emmeche et al. 2000; El-Hani, Videira 2001; Hulswit 2005). In other words, strong DC demands an acceptance of substance dualism, and, thus, it is blatantly incompatible with a scientific understanding of emergence. Moreover, this notion faces a number of other important difficulties. If we consider the standard case in discussions about DC, i.e., “reflexive” and “synchronic” downward causation (Kim 1999), in which some activity or event involving a whole at a time t is a cause of, or has a causal influence on, the events involving its own micro-constituents at that same time t, then a strong account of DC looks like a bizarre metaphysical bootstrapping exercise (see Symons 2002). Therefore, there are only two viable candidates for a scientifically acceptable account, both committed to an interpretation of DC as a case of synchronic formal causation: medium and weak DC. We can summarize the key points in the arguments for medium DC (cf. Emmeche et al. 2000) as follows: (i) a higher-level entity comes into being through the realization of one amongst several possible lower-level states. (ii) In this process, the previous states of the higher level operate as “factors of selection” for the lower-level states. (iii) The idea of a factor of selection can be made more precise by employing the concept of “boundary conditions”, introduced by Polanyi (1968) in the context of biology, particularly in the sense that higher-level entities are boundary conditions for the activity of lower levels, constraining which higher-level phenomenon will result from a given lower-level state. (iv) Constraints can be interpreted in terms of the characterization of a higher level by “organizational principles” — law-like regularities — that have a downward effect on the distribution of lower-level events and substances. (v) Medium DC is committed to the thesis of “constitutive irreductionism”, namely, the idea that even though higher-level systems are ontologically constituted by lower-level entities, the higher level cannot be reduced to the form or organization of the constituents. (vi) Rather, the higher level must be said to “constitute its own substance” and not merely to consist of its lower-level constituents, or, else, a higher-level entity should be regarded as a “real substantial phenomenon” in its own
Chapter 7. Emergence of semiosis: a general model
157
right. (vii) This interpretation of DC may assume either a thesis Emmeche et al. (2000) call “formal realism of levels”, stating that the structure, organization or form of an entity is an objectively existent feature of it which is irreducible to lower-level forms or substances, or a thesis they designate as “substantial realism of levels”, claiming that a higher-level entity is defined by a “substantial difference” from lower-level entities. The difference from strong DC is said to lie in the necessary commitment, in this position, to the thesis of a “substantial realism of levels”. In turn, the idea of weak DC can be summarized in terms of the following arguments: (i) in the weak version, DC is interpreted in terms of a “formal realism of levels”, as explained above, and “constitutive reductionism”, the idea that a higher-level entity ontologically consists of lower-level entities organized in a certain way. (ii) Higherlevel forms or organization are irreducible to the lower level, but the higher-level is not a “real substantial phenomenon”, i.e., it does not add any substance to the entities at the lower level. (iii) In contrast to the medium version, weak DC does not admit the interpretation of boundary conditions as constraints. (iv) By employing phase-space terminology, Emmeche and colleagues explain weak DC as the conception of higher-level entities as attractors for the dynamics of lower levels. Accordingly, the higher level is thought of as being characterized by formal causes of the self-organization of constituents at a lower level. (v) The relative stability of an attractor is taken to be identical to the downward “governing” of lower-level entities, i.e., the attractor functions as a “whole” at a higher level affecting the processes that constitute it. (vii) The attractor also functions as a whole in another sense of the word, given that it is a general type, of which the single phase-space points in its basin are tokens. The contribution in Emmeche et al. (2000) to the debates about DC stressed a diversity of DC accounts that has been often neglected, and, moreover, tried to make some advance in organizing the variety of such accounts. As Hulswit (2005) sums up, it was “[…] a valiant attempt at creating some order in the conceptual chaos that characterizes the discussion regarding downward causation”. Nevertheless, as we see it now, its typology faces a number of problems. In particular, the distinctions between strong, medium, and weak DC should be further clarified. For instance, it seems necessary to describe in more detail in what sense strong and medium DC differ as regards
158
Genes, Information, and Semiosis
the idea that a higher-level entity is a “substantial” phenomenon, or, else, how one would differentiate medium versions committed to the thesis of a “substantial realism of levels” from strong DC. For the sake of our arguments, we will simply work below with an interpretation which comes close to medium DC by interpreting boundary conditions as constraints, but, at the same time, departs from it, by resolutely rejecting “constitutive irreductionism”. It also comes close, thus, to weak DC. However, it is not our main purpose to classify our account, but rather to explain how we conceive the relationship between DC and constraints. In order to do so, we will begin by considering that, when lowerlevel entities are composing a higher-level system, the set of possible relations among them is constrained, since the system causes its components to have a much more ordered distribution in spacetime than they would have in its absence. This is true in the case of both entities and processes, since processes also make the elements involved in them assume a particular distribution in spacetime. We can take a first step, then, towards explaining why the same lowerlevel entity can show different behaviors depending on the higherlevel system it is part of — the basis for a concept of irreducibility based on the non-deducibility of the components’ behavior. We can plausibly argue that lower-level entities are ‘enslaved’ by a particular pattern of constraints established by the higher-level structure in which they are embedded, so that their relations to each other are modified, and, consequently, their causal dispositions. The research field of non-linear dynamics is rich with examples of this phenomenon (as shown, for instance, by the work of Alwyn Scott; cf. Scott 2004). The ‘causes’ in DC can be treated, in these terms, as higher-level general organizational principles which constrain particular lowerlevel processes (the ‘effect’), given that the particular relations the parts of a system of a given kind can be engaged in depend on how the system’s structures and processes are organized. In this framework, DC can be interpreted as a ‘formal cause’ by recasting the notion of higher-level ‘constraints’ (or ‘constraining conditions’), much discussed in works about the nature of complex systems (e. g. Salthe 1985), in terms of Aristotle’s set of causal concepts (see Emmeche et
Chapter 7. Emergence of semiosis: a general model
159
al. 2000; El-Hani, Pereira 2000; El-Hani, Emmeche 2000; El-Hani, Videira 2001).84 The notion of ‘boundary conditions’, introduced by Polanyi (1968), is useful for characterizing these higher-level constraints (see also van Gulick 1993). Polanyi argued that a living system, as a naturally designed entity, works under the control of two principles: The higher one is the principle of design or organization of the system, and this harnesses the lower one, which consists in the physicochemical processes on which the system relies. As the physicochemical processes at the lower level are harnessed, the components come to perform functions contributing to the maintenance of the dynamical stability of the system as a whole. Hulswit (2005) recently argued that the meaning ascribed to the term ‘causation’ in debates about DC usually refers to ideas closer to ‘explanation’ and ‘determination’ than to ‘causation’, provided we understand causation in the intuitive sense of ‘bringing about’, i.e., in the sense of efficient causation. Not surprisingly, he considers the expression ‘downward causation’ badly chosen. Although verbs usually related to the causing activity of a higher level in DC, such as ‘to restrain’, ‘to select’, ‘to organize’, ‘to structure’, ‘to determine’ etc., may be understood as being related to ‘causing’ (in the sense of ‘bringing about’), they are certainly not equivalent to ‘causing’ (Hulswit 2005). This can be seen as a result of an impoverishment of the meaning of the term ‘cause’ in modern science, due to the fact that classical physics critically appraised, and, ultimately, denied a number of theses related to Aristotelian philosophy, many of them concerned with the principle of causality (ElHani, Videira 2001). Ultimately, only two of the four Aristotelian 84 Notice that a case can be made for sticking to the vocabulary of constraints without introducing a potential source of contention such as the Aristotelian formal causal mode. We will come back to this possibility later. For the moment, consider that it was quite natural that, as the problem we are dealing with concerns causation, it seemed worth exploring, for a number of authors, the consequences of recasting the treatment of constraining conditions in terms of formal causal influences of wholes over parts. Nevertheless, we should not lose from sight that the very difficulties faced by the concept of DC, which are indeed the main motivation for emergentist thinkers to seek new ways of understanding causality, can be seen as an evidence that causality is the wrong issue when comes to emergence (at least in some domains. See Pihlström 2002) and, generally speaking, complex systems (see Van de Vijver et al. 2003).
160
Genes, Information, and Semiosis
causal modes, efficient and final causes, came to be included in the meaning ascribed to the term ‘cause’ in most modern languages. Symptomatically, the Greek word translated as ‘cause’ (archai) in Aristotle’s works does not mean ‘cause’ in the modern sense (Ross, [1923]1995: 75; Lear 1988: 15).85 For Aristotle, a ‘cause’ was not only an antecedent event sufficient to produce an effect or the goal of a given action, but the basis or ground of something. In other terms, to refer to Aristotle’s archai as ‘causes’ can be very misleading; it is arguably better to treat them as ‘principles’. It is in this sense that Aristotle can conceive matter and form as also having the nature of ‘causal’ modes — in terms of his material and formal causal modes. It is not surprising, then, that, if we stick to our currently intuitive ideas about causation, Aristotle’s causal modes seem more similar to modes of explanation than to modes of causation. (And the same is true of interpretations of DC which appeal to ideas such as that of formal causes). Aristotle seemed to be thinking mainly about the grounds for our understanding, while pondering about causal modes. It would be possible to use the above claims as a ground for counteracting Hulswit’s arguments. But we do think he pointed out important limitations in recent accounts of DC, and we will rather employ these claims as a basis for combining Hulswit’s ideas with some tenets advanced by those accounts. Therefore, we have been involved in inquiring further into his observation that, although verbs usually related to the causing activity of a higher level in DC, such as ‘to restrain’, ‘to select’, ‘to organize’, ‘to structure’, ‘to determine’, etc., may be understood as being related to ‘causing’, they are not equivalent to ‘causing’, in the modern sense (El-Hani, Queiroz 2005b). But how should we understand the relationship between these higher-level ‘activities’ and ‘causing’? It seems to us that the important relation between the ideas usually connected to DC in recent accounts and the basic ideas involved in causation concerns the fact that, in both cases, we are dealing with some kind of determination. As Hulswit (2005) stresses, the main difference between ‘determining’ and ‘causing’ is that the former primarily involves necessitation (in the sense of ‘it could not be otherwise’ or, to put it in terms more consistent with probabilistic 85
Translated into Latin, archai turned into causae, which in turn was translated into English as ‘cause’ (and, equivalently, in the case of other languages).
Chapter 7. Emergence of semiosis: a general model
161
events, ‘it would not tend to be otherwise’) while the latter primarily involves the idea of ‘bringing about’. We would like to invite our readers, then, to consider three issues: first, that most of the debates about DC are really about determination or explanation rather than causation; second, that efficient causes are typically regarded as particulars (usually events, facts, or substances), and ‘downward causes’ are more properly interpreted (in our view) as general, law-like organizational principles86; and, third, that a similar appeal to ‘determination’ has been made in the case of another determinative but mereological relation, namely, physical realization (and, consequently, supervenience), that cannot be properly accounted for as ‘causal’ (see Kim 1993, 1996). We will be able to see, then, how proper it is to advance the claim that it is better to refer to downward (formal) determination, rather than downward causation: […] so-called ‘downward formal causation’ is neither a species of downward causation, nor of downward explanation, but […] it is first and foremost a species of downward determination. (Hulswit 2005. Emphases in the original)
Instead of proposing that an understanding of the influence of wholes over parts demands causal categories other than efficient causation, we can rather claim that such understanding requires other kinds of determination than just causation. In fact, causes are not the only sort of determining factors in the world and, in fact, it is largely accepted in other current philosophical debates, such as those about supervenience, the introduction of non-causal determinative relations.87 Given the arguments presented above, we can reformulate the ninth question as follows: (9) can we describe any sort of downward determinative relation in semiosis?
86 Precisely for this reason, Campbell (1974) himself remarked that the expression ‘downward causation’, which he first explicitly used, was ‘awkward’. 87 As much as in the case of DC, a proper explanation of downward determination demands a clear theory about the relata at stake and the connections between this kind of determination and other basic categories, such as “law” and “cause” itself. Here, we will not expand on these issues. For a discussion about the first one, see El-Hani, Queiroz (2005b).
162
Genes, Information, and Semiosis
Unpredictability It is now time to turn to diachronic emergentism, which can be treated rather briefly here. This variety of emergentism is concerned with the doctrine of ‘emergent evolution’. All diachronic theories of emergence are ultimately grounded on the thesis that ‘novelties’ occur in evolution, opposing any sort of preformationist position. But merely the addition of the thesis of novelty does not turn a weak emergence theory into a strong one. Strong forms of diachronic emergentism demand the thesis of ‘in principle theoretical unpredictability’ of novel properties or structures.88 The notion of ‘genuine novelty’ then enters the scene, as one claims that a given property or structure is not only novel but also could not be theoretically predicted before its first appearance. A systemic property can be unpredictable in this sense for two different reasons (Stephan 1998: 645): (i) because the microstructure of the system exemplifying it for the first time in evolution is unpredictable; (ii) because it is irreducible, and, in this case, it does not matter if the system’s microstructure is predictable or not.89 As the second case does not offer any additional gains beyond those obtained in the treatment of irreducibility, we will focus our discussion on the unpredictability of the structures of semiotic systems and processes. We can reformulate, then, the eighth question raised in the previous section as follows: (8) is the structure of semiotic systems or processes in principle theoretically unpredictable? Now, we should turn to our answers to the questions raised along the previous sections. 88
Notice that theoretically-unpredictable structures or properties can be inductively-predictable (Kim 1999), given that, once a structure or property appears for the first time, it is possible that further occurrences of that structure or property are adequately predicted, given the thesis of synchronic determination. Moreover, “in principle” unpredictability is introduced in opposition to “practical” unpredictability, which is dependent on our cognitive limitations and state of knowledge. 89 Notice that the two reasons for the unpredictability of systemic properties have very different status. While the second is empirical in nature, particularly if irreducibility is interpreted in terms of non-deducibility, the former depends, as we will see in the argument below that semiosis is in principle unpredictable in the Peircean framework, on specific metaphysical commitments that a thinker can clearly reject.
Chapter 7. Emergence of semiosis: a general model
163
2.3. Answering the questions about semiosis What is a semiotic system? Let us consider, first, the following question: (1) what is a semiotic system? In previous chapters, we employed this expression without defining it. It is time to introduce a definition. But, first of all, we should offer a definition of ‘system’, in more general terms. A system is usually defined as a set of elements that maintain relations with one another (Bertalanffy 1973: 55; Pessoa Jr. 1996: 30). By ‘elements’ we mean primitive entities which are found at each instant in one among several possible states. Elements establish ‘relations’ when the state of an element depends on the state of another one. Some definitions of ‘system’ include other items, such as Bunge’s (1977) definition, in which a system x is defined by its composition — the set of its components —, structure — the set of relations among its components —, and environment — the set of other systems with which x establishes relations. A significantly related but slightly more refined way of defining systems is found in dynamical systems theory, in which systems are conceived as sets of interdependent variables. By ‘variable’ we mean some entity that can change, i.e., that can be in different states at different times — it is obvious that the concepts of “variable” and “elements”, as stated here, are quite similar. The state of a system is simply the state or value of all its variables at a given time t. The behavior of a system, in turn, consists of transitions between states (Gelder 1998: 616). Now we can turn to a definition of what is a “semiotic system”. Fetzer (1988) called ‘semiotic system’ a system that produces, transmits, receives, and interprets Signs of different kinds. Fetzer argues that what makes a system ‘semiotic’ is the fact that its behavior is […] causally affected by the presence of a sign because that sign stands for something else iconically, indexically, or symbolically, for that system. Those things for which signs stand, moreover, may include abstract, theoretical, nonobservable, or non-existent objects and properties, which may be incapable of exerting any causal influence on a system themselves. (Fetzer 1997: 358)
Semiosis can be defined as a self-corrective process involving cooperative interaction between three components, S-O-I. Therefore,
164
Genes, Information, and Semiosis
as a straightforward consequence of the nature of semiosis, semiotic systems show self-corrective behavior, or some kind of goal-directed activity (see Ransdell 1977: 162). Such a self-corrective behavior depends on the capability of semiotic systems of using Signs as media for the communication of forms from Objects to Interpretants. Peirce suggested that semiotic systems could be treated as the embodiment of semiotic processes (see CP 5.314). Surely, this blurs a distinction between entities and processes which characterized Western thinking for most of its history. Peirce was, as mentioned above, a process thinker, a representative of a philosophical tendency of treating processes as being more fundamental than entities as ontological categories (see Rescher 1996). To give ontological primacy to coordinated, organized family of occurrences clearly contradicts the priority historically given to entities in most of the Western thinking, substantially influenced by Aristotelian philosophy. But such a disagreement can be seen as part of a criticism of the ‘substance paradigm’ or ‘myth of the substance’ (Seibt 1996), a criticism that has been put forward by a number of thinkers, such as Alfred N. Whitehead, Charles S. Peirce, Charles Hartshorne, Paul Weiss, Samuel Alexander, Conway Lloyd Morgan, and Andrew Paul Ushenko (see Rescher 1996, 2002). It is not that process philosophy should necessarily claim that the idea of entities has to be abandoned. It is only that, when considering entities, we should always bear in mind that processes should be treated, in a dynamical world, as more fundamental than the former, since “… substantial things emerge in and from the world’s course of changes” (Rescher 1996: 28). Or, to put it differently, a process philosophy can address entities, as Peirce does in the case of semiotic systems, as relatively stable bunches of processes, which emerge from processes and subsequently vanish into processes. A semiotic system can be understood in these terms as a relatively stable (both spatially and temporally) cluster of semiotic processes.
Are semiotic systems physically constituted? A second question concerns the nature of semiotic systems: (2) are they physically constituted? Sign processes are relationally extended within the spatiotemporal dimension, so that something physical has
Chapter 7. Emergence of semiosis: a general model
165
to instantiate or realize them. This means that Signs cannot act unless they are spatiotemporally realized (see Emmeche 2003; Ransdell 1977). Therefore, semiotic systems should be physically embodied (Emmeche 2003; Deacon 1999). If a Sign is to have any active mode of being, it must be physically instantiated. In this connection, it is important to remember that Peirce considered the material qualities of the Sign as the characters that belonged to the Sign in itself: “Since a sign is not identical with the thing signified, but differs from the latter in some respects, it must plainly have some characters which belong to it in itself, and have nothing to do with its representative function. These I call the material qualities of the sign” (CP 5.287).90
Are semiotic systems new? A third question asks (3) whether semiotic systems can be regarded as forming a new class of systems, with new structures, instantiating new properties, processes, behaviors, dispositions, etc. It is not our problem here to define where is the threshold beyond which semiotic systems are found in the history of the universe. We assume, for the sake of our arguments, that there was a period in which systems capable of using Signs did not exist. Therefore, even though irreducible triadic relations may have preceded the origins of semiotic systems, we postulate that this class of systems arose in the course of evolution. We consider, then, that before the emergence of semiotic systems, only reactive systems existed, which were not capable of interpreting, and, thus, using Signs. Surely, there were things in the world to which physically embodied natural systems reacted, but these systems were not able to use Signs as media for the communication of forms, i.e., they were not interpreters. Nothing but a dynamics of systems and things dyadically coupled existed, with no interpretative processes taking place. Given this reasonable set of assumptions, we 90
Notice that, even if one assumes that there can be semiosis before the emergence of semiotic systems, the Sign processes at stake would still have to be physically instantiated or realized in one way or another, since the attribution to Sign processes of the property of being relationally extended within the spatiotemporal dimension does not depend on whether semiotic systems are involved or not.
166
Genes, Information, and Semiosis
can say that semiotic systems are a new class of systems, with a new kind of structure, capable of producing and interpreting Signs, and, thus, of realizing semiosis, as a new kind of (emergent) process. The emergence of the competence to handle Signs changed the dynamics of the evolution of natural systems. After all, we can claim that semiotic systems show modes of evolution not found among merely reactive systems. For instance, living systems which possess Signs in the form of DNA can evolve by a process in which past successful interactions between a system and its environment are represented in Signs which are passed over to the next generations, influencing the future evolution of the lineage to which the system pertains, as discussed in Ch. 6. Furthermore, after the competence to handle Signs, and, thus, instantiate semiosis emerged, the evolution of semiotic systems did not cease, but, rather, new kinds of such systems emerged, operating with different classes of Signs (e. g., iconic, indexical, symbolic) and evolving in different manners (see Fetzer 1988, 1997; Queiroz, El-Hani 2006b). At first, the idea that semiotic systems constitute a new class of systems seems to be incompatible with a basic feature of Peirce’s metaphysical framework, namely, synechism. After all, the doctrine of emergence is committed to the idea that the evolution of the universe shows discontinuities, and synechism is a “tendency to regard everything as continuous” (CP 7.565). According to Peirce (CP 6.169), synechism is “… that tendency of philosophical thought which insists upon the idea of continuity as of prime importance in philosophy and, in particular, upon the necessity of hypotheses involving true continuity.”91 We claim, however, that this incompatibility is only apparent, since an emergentist philosophy can be seen as providing a way of overcoming the dichotomy between continuity and discontinuity. Such emergentist philosophy can accommodate, in our view, Peirce’s synechism. For instance, Morgan’s (1923) Emergent Evolution, regarded by Blitz (1992) as the founding work in the tradition of emergentism, provides an emergence theory that combines the ideas of continuity and discontinuity. Among the fundamental theses of Morgan’s theory of emergent evolution, we find two which are directly consequential to our present 91
For further discussions about synechism, see Parker (1998), Potter (1997), Murphey (1993).
Chapter 7. Emergence of semiosis: a general model
167
discussion: the theses of the co-occurrence of emergents and resultants, and of quantitative continuity and qualitative novelty.92 For Morgan, emergent properties were never instantiated at a given level without occurring along with resultant properties, which could be predicted on the grounds of theoretical knowledge about the previous level and conferred continuity to the evolutionary process. Thus, even though emergence concerns the appearance of genuinely new properties that could not be predicted from knowledge about preexistent entities described at a lower level, it does not amount in Morgan’s theory to a gap in the evolutionary process. Therefore, it is not in the sense of some sort of leap in evolution, which would be indeed incompatible with synechism, that Morgan put forward the claim of qualitative novelty in evolution. Rather, he conceived qualitative novelty in terms of a qualitative change of direction or a critical turning point in an otherwise continuous evolutionary process. In Morgan’s (1923: 5) own words, “… through resultants there is continuity in progress; through emergence there is progress in continuity.” Consider, also, that it is the very process of gradual and quantitative change of natural systems which creates, in Morgan’s framework, the conditions for the qualitative change related to the notion of emergence. This qualitative change, in turn, has the character of a critical turning point (or, as physicists might prefer, a kind of phase transition) because it establishes new kinds of relatedness among pre-existent entities or events, and, thus, changes the mode of evolution of natural systems. It is clear, then, that emergence is related to punctuations in a continuous process, rather than to a mere jump in evolution. A number of quotations from Morgan’s seminal work on emergence will suffice to show that property emergence is related to critical turning points in which new patterns of organization (and, thus, constraints) are established in the evolution of systems. Morgan characterizes ‘emergent evolution’ as follows: Evolution, in the broad sense of the word, is the name we give to the comprehensive plan of sequence in all natural events. But the orderly sequence, historically viewed, appears to present, from time to time, something genuinely new. Under what I here call emergent evolution stress is laid on this incoming of the new. (Morgan 1923: 1) 92
See Morgan (1923), Blitz (1992). The theses were named by Blitz.
168
Genes, Information, and Semiosis
He also states that “... the emergent step [...] is best regarded as a qualitative change of direction, or critical turning point, in the course of events” (Morgan 1923: 5), connecting emergent events to the “... expression of some new kind of relatedness among pre-existent events” (Morgan 1923: 6). Finally, Morgan observes: When some new kind of relatedness is supervenient (say at the level of life), the way in which physical events which are involved run their course is different in virtue of its presence — different from what it would have been if life had been absent. [...]. I shall say that this new manner in which lower events happen — this touch of novelty in evolutionary advance — depends on the new kind of relatedness which is expressed in that which Mr. Alexander speaks of as an emergent quality. (Morgan 1923: 16. Emphases in the original)
These quotations from Morgan’s classical work on emergence contain some basic criteria for treating properties and processes as ‘emergent’. First, they should be (genuinely) new under the sun. By qualifying the novelty as ‘genuine’, we highlight a requirement stressed by many emergentists, namely, that emergent properties and processes should not only be ‘new’, but also ‘unpredictable’ (see above). Second, they should be closely connected with the appearance of a new kind of relatedness (and, thus, of a new organizational principle) among preexistent processes and entities, entailing a modification in the way lower-level events run their course, and, consequently, some sort of downward determination. Therefore, novelty in emergent evolution remarkably appears in the form of novel organizational principles, so that a primacy is given to the emergence of structured processes and entities. Third, the emergence of properties or processes in a new class of systems (as defined by the above-mentioned new kind of relatedness) should change the mode of systems’ evolution. This change, in turn, should be precisely the result of a modification of the behavior of pre-existent entities and processes under the influence of that new kind of relatedness (and this again leads us to the issue of downward determination). It is clear from this exposition that Morgan’s theory does not postulate jumps in the evolutionary process. Therefore, a commitment to an emergence theory doesn’t lead to a necessary contradiction in the proposal of a joint emergentist and Peircean account, despite the central role of synechism in Peirce’s thought. Rather, any emergentist philosophy that subscribes to a rendering of the relationship between
Chapter 7. Emergence of semiosis: a general model
169
continuity and novelty along the lines developed by Morgan should explicitly claim that resultant properties provide a quantitative continuity in evolution, upon which qualitative novelties arise from time to time as changes in the direction of evolution, rather than as saltationist leaps. We hope these arguments are enough to show that there is no necessary contradiction between Peirce’s doctrine of synechism and emergentist philosophy. We can speculate that the competence to handle Signs appeared in the evolution of systems as the product of a continuous process. Nevertheless, when semiotic systems appeared, they exhibited a way of behaving which was significantly different from that of reactive systems, as they could go beyond a mere coupling to their circumstances, being able to interpret them. It is reasonable to suppose, then, that that difference in behavior entailed a distinct mode of evolution in the case of semiotic systems, as compared to reactive systems. Thus, we can hold that a qualitative change, a critical turn in evolution, took place with the appearance of semiotic systems. After all, a system which is capable of interpreting the world through the mediation of Signs evolves in a manner which is determined by the fact that they are capable of using Signs to obtain information about the environment in such a way that those Signs perform functions favoring their survival and/or reproduction (Emmeche 1997, 1998).
Is semiosis a systemic process? We should turn now to our fourth question: (4) can semiosis be regarded as a systemic process? Consider, first, that according to the model developed above the actualization of potential chains of triads depends on boundary conditions established by a macro-semiotic level amounting to networks of chains of triads. It is possible to conceive of the macro-semiotic level as corresponding to the whole semiotic system, based on the idea that the latter can be treated as the embodiment of semiotic processes (see above). Therefore, although semiosis is instantiated at the focal level, it should be understood as a systemic process, given that the macro-semiotic level establishes the boundary conditions required for its actualization. To put it differently, the very instantiation of semiosis at the focal level depends on
170
Genes, Information, and Semiosis
a constraining influence of the semiotic system as a whole (i.e., the macro-semiotic level). A variety of theoretical tools can be used to build a more detailed mapping of these relations. The interdisciplinary field between semiotics and systems theory, still only sparsely investigated, is a promising area for the search of new principles of emergent signifying complexity. The model we have proposed here is just one of several possible approaches (see also, e. g., Cariani 1989; Pattee 2000; Joslyn 2001). As to the fifth question — (5) How should we describe the levels in a semiotic system —, the model presented in section 1 can be seen, as a whole, as an answer to it.
Is semiosis synchronically determined by the properties and arrangement of the parts in a semiotic system? Sixthly, we asked: (6) in what sense can we say that semiosis, as an emergent process, is synchronically determined by the properties and arrangement of the parts in a semiotic system? In our hierarchical model, semiosis is located at the focal level, instantiated as chains of triads, while individual triads are situated at the lower level, and networks of chains of triads, at the higher level. Therefore, while considering the idea of synchronic determination, we have to focus our attention on the relationship between chains of triads, at the focal level, and individual triads, at the micro-semiotic level. Semiosis is described by Peirce as a pattern of determinative relationships between functionally specified correlates. We consider, here, that this description entails the idea that semiosis is synchronically determined by the microstructure of the individual triads composing a chain of triads, i.e., by the relational properties and arrangement of the elements S, O, and I.93 There cannot be any 93
To understand our argument in a clear way, it is very important to avoid conflating synchronic with diachronic determination. We claim here that a Peircean framework accommodates a thesis of synchronic determination, while denying any claim of diachronic determination. Another source of conflation is to take the notion of determination (in synchronic determination) as equal to or entailing causal determinism. First, according to the semiotics and metaphysics of Peirce, deterministic causality does not reign completely due to the real vagueness and spontaneity intrinsic in natural processes. Second, as we argued above, “determination” and “causation” should not be confused.
Chapter 7. Emergence of semiosis: a general model
171
difference in semiosis without a difference in the properties and/or arrangement of S, O, and I. The arrangement of the elements S-O-I is specified by the triadic relations of determination between them. Otherwise, it would be a mere juxtaposition of three elements (see CP 1.371, 1.363; see Brunning 1997). The properties of S, O, and I are relational because these elements are engaged in irreducibly triadic ordered relations. As Savan 1988: 43) writes, “the terms interpretant, sign and object are a triad whose definitions are circular. Each of the three is defined in terms of the other two.” In fact, the only property of S, O, and I, as functional roles, is to be in a specific position in an irreducible triadic relation to one another, namely, to be the first, the second, or the third terms in such a relation (see Tienne 1992). One should also consider the modal strength of the relation of synchronic determination between chains of triads and triads. We will consider here four standard possibilities (see Bailey 1999): (i) weak necessity, in which the determinative relation holds in the actual world, but need not hold in any other possible world; (ii) Natural, or physical, or nomic, or nomological necessity, in which the determinative relation holds in the actual world and in all naturally possible worlds, which can be described, very roughly, as all worlds in which the physical laws sufficiently resemble actual laws; (iii) Metaphysical necessity, in which the determinative relation holds in the actual world and in all metaphysically possible worlds, which comprise all worlds where a posteriori necessary truths (such as “water is H2O”) hold; and (iv) Logical necessity, in which the determinative relation holds in the actual world and all logically possible worlds, roughly, those where a priori necessary truths hold — this is the set of all possible worlds. In the case of semiosis, we propose that the determinative relations between the elements of individual triads, as well as between triads, in a chain of triads, hold with logical necessity.94 Initially, consider that the demonstration that S-O-I constitute an indecomposable relation 94
Notice that while discussing the logical relations between elements and triads, we are working in the domain of Speculative Grammar, the study of the “general conditions of signs being signs” (CP 1.444). For Houser (1997: 9), “the logician who concentrates on speculative grammar investigates representation relations (signs), seeks to work out the necessary and sufficient conditions for representing, and classifies the different possible kinds of representation.”
172
Genes, Information, and Semiosis
should be first carried out logically (see Houser 1997: 16). The reason for the precedence of a formal treatment of relations over the empirical and metaphysical treatments lies in the fact that only formally one can perform an analysis of the properties of completeness and sufficiency of Peirce’s categories (Parker 1998: 3, 43). It is only subsequently that the property of logical irreducibility should be checked in the empirical and metaphysical domains. The precedence of the logical treatment has methodological consequences. An analysis of the formal properties, in contrast with the material properties95, should precede any empirical or metaphysical investigation of the categories. In other words, a logical-mathematical analysis of the categories should be previous to any formulation in the domains of phenomenology, normative sciences and metaphysics96, which employ mathematical techniques and results to validate the categories established by the logical treatment of relations (see Hookway 1985: 182; Parker 1998). Therefore, in our discussion about the modal strength of the relation of synchronic determination between chains of triads and triads, we will begin with a logical treatment of the relations between the elements of semiosis. We will focus our attention, first, on the functional roles of S, O and I, as established in a logical analysis of their relations. The functional roles of S, O and I are logically determined in each triad, as regards both the relationships within a triad and the constitution of chains of triads. Therefore, these determinative relations hold with logical necessity: in a world substantially different from the actual world in its physical laws, i.e., a world nomologically distinct from the actual world, the logical relationships between S, O and I would still be the same. If we are right in our arguments, then these relations hold in the set of all possible worlds, provided that the conceived world allows the existence of physical entities or processes. After all, there is an important constraint for something to be a semiotic system, namely, that it should be physically embodied (see above). This does not mean that the determinative relations between S, O, and I in a semiotic process might be only nomologically valid, but rather that any 95
The distinction between material and formal properties was clearly established by Peirce after 1885 (see Kent 1997: 448). 96 For an introduction on phenomenology, normative sciences, and metaphysics, see Waal (2001), Parker (1998).
Chapter 7. Emergence of semiosis: a general model
173
logically conceivable world in which semiosis can take place is a world in which natural laws allow the existence of physical entities or processes, which are a necessary condition for semiosis. In any such world, then, the determinative relations between S, O and I hold with logical necessity. If we suppose, for the sake of our argument, that there are logically conceivable worlds where no physical entities or processes are present, it will be simply the case that such worlds will not show any semiotic process or system, and, thus, no determinative relation at all between the elements involved in semiosis will take place there. In the empirical domain, in turn, we should focus our attention not only on the functional roles of S, O and I, but also on how these functional roles may be embodied, and how the relations between them may be instantiated in the actual world. In this case, notice that while the functional roles are logically determined, the occupants of the functional roles of S, O and I are contingent. For instance, that the word ‘elephant’ is a Sign for that big animal in the world can be treated as a contingent fact; that is, it is not logically necessary that the word ‘elephant’, as an occupant of the functional role of a Sign (S), stand through the Interpretant (I) for that big animal, the occupant of the functional role of the Object (O) in the example at stake. But the determinative relationships between these elements are logically determined, and, consequently, are also the functional roles of S, O, and I. Thus, in a world sufficiently distinct from the actual world in its physical laws, entirely different entities or processes might be playing the functional roles of S, O and I in distinct semiotic systems. We can conclude that the fact that a given class of entities or processes plays a functional role in a semiotic process holds with nomological rather than logical necessity, even though the functional role itself holds with logical necessity. Now, where does this take us? As to the sixth question — (6) in what sense can we say that semiosis, as an emergent process, is synchronically determined by the properties and arrangement of the parts in a semiotic system — we can approach an answer with some precautions. The answer is that yes, there is a synchronic determination, holding with nomological rather than logical necessity, of semiosis as an emergent process, as indicated above, to the extent that this conception does not conflict with the following precautionary remarks, that we make to emphasize the special nature of semiotic
174
Genes, Information, and Semiosis
processes as compared to conceptualizations of processes that typically appear in more traditional forms of systems thinking (as the latter to some extent has been influenced by the aforementioned Western metaphysical paradigm of substance thinking). (a) Habitualness: The determinative relation here has to be conceived within the Peircean framework as not completely deterministic, but including probabilistic tendencies and, thus, even ever so slight deviations from the lawfulness inherent in the various modes of development and evolution, i.e., not necessitarianism, but Thirdness, as seen in the formation of habits, new codes, and new rules of interpretation. (b) Generativity: The relation of synchronic determination should not be in conceptual conflict with the idea of generative semiosis in interpretative relations, that is, the dynamic generation of new Interpretants. (c) Interdependency: The ‘parts’ and ‘wholes’ of the semiotic system must be conceived of as not discrete and separate substantial parts, but as semiotic relations which are context-dependent and where the whole is partly specifying the semiotic function of the parts (whether that expansion of the notion of synchronic determination dramatically changes that concept into another is a possible concern we shall not pursue in detail here). (d) Dynamism: The analytical philosophical notion of ‘synchronicity’ (as in the term ‘synchronic determination’) should here be conceived within a Peircean synechistic perspective, in which ‘being synchronously determined’ does not preclude the existence of a processual finality and purpose-directiveness in natural semiotic processes, and in which it implies a view of temporality not as a spatialized line of single points representing atomic disconnected instances, but rather a synechistic view of temporality, including a future-directedness also found in interpretative processes. These precautionary remarks about the special nature of semiotic systems may indicate some problems in reconciling a systems view on emergence (mainly derived from contemporary natural science and metaphysical discourse as influenced by analytical philosophy) with a Peircean semiotic view on sign production, mediation and interpretation. It is an important philosophical task to resolve such real or apparent conflicts, but we will not delve on these problems here, since they cannot be dealt with briefly, and in the end, they may not in this context be crucial for our major arguments.
Chapter 7. Emergence of semiosis: a general model
175
In what sense is semiosis irreducible? Among the several properties related to semiosis (processuality, CP 5.484; irreversibility, CP 5.253, 5.421; continuity, MS 875, see also Parker 1998: 147; tendency to the infinitum, CP 2.92, 2.303; vagueness, CP 5.447; generality, CP 6.172, see Potter 1997: 89; regularity; growth; lawfulness), we can say that the relational irreducibility of the triad is one of the most, if not the most, important. Thus, the next question (7) is particularly important, as it concerns the interpretation of the principle of the irreducibility of semiosis. The semiotic triadic relation is regarded by Peirce as irreducible in the sense that it is not decomposable into any simpler relation (e. g., CP 5.484). As Peirce carefully discusses the irreducibility of triads, we will consider in the following arguments what we defined above as the micro-semiotic level. We argue, first, that the semiotic relation is not irreducible because the condition of analyzability is violated. Peirce would accept, in our view, that the properties a triad possesses must follow from the behavior of its elements, including the very property of being semiotic. If we know the relations in which any three elements are involved, then we will be able to know also whether the process in which they are engaged is semiotic, since we will know whether or not the elements play the logical-functional roles of S, O, and I. To put it differently, non-analyzability or what Boogerd et al. (2005) call the vertical condition for emergence is not the reason why we should consider, in a Peircean framework, semiosis as an irreducibly triadic relation. We can understand why a semiotic relation is irreducible, in a Peircean framework, on the grounds of the second notion of irreducibility discussed above, based on the non-deducibility of the behavior of the system’s parts. In this case, we should show that the specific behavior of the elements of a triad does not follow from the elements’ behaviors in simpler relations. We think that semiosis can even be regarded as the best example of a class of relations in which the second condition for reducibility discussed above is violated, since the behavior of the elements of a semiotic relation does not follow from the behavior they show in isolation or in dyadic relations. The functional roles of the elements in a semiotic relation cannot be identified in structures simpler than a triadic relation. The functional role of S can be identified only in the mediative relation it
176
Genes, Information, and Semiosis
establishes between O and I. Similarly, the functional role of O is identified in the relation by means of which it determines I through the mediation of S. And, finally, the functional role of I is identified by the fact that it is determined by O through S. Therefore, if we consider only dyadic relations, S-I, S-O, or I-O, or the elements of a triad in isolation, we cannot deduce how they would behave in a triadic relation, S-O-I (EP 2:391). Therefore, the irreducibility of semiosis should be understood in terms of the non-deducibility of the behavior of the logical-functional elements of a triad on the grounds of their behavior in simpler relations. Or, to put it differently, rather than the vertical, the horizontal condition for emergence (Boogerd et al. 2005) holds in the case of semiosis.97 Is downward determination involved in semiosis? We will turn, now, to another question about the understanding of semiosis as an emergent process: (9) can we describe any sort of downward determinative relation in semiosis? If we consider, first, the relationship between the macro-semiotic level and semiosis at the focal level, we can argue that it involves a determinative downward relation. More specifically, as the model presented above shows, downward determination (DD) in semiotic phenomena can be conceptualized as boundary conditions which select, among the potentialities established by the micro-semiotic level, those semiotic processes which will be actualized at a given time t. This determining influence can be conceptualized as the selection by a higher-level general organization principle, observed in a macro-semiotic network, on semiosis, as a particular lower-level process. Nevertheless, as triads and chains of triads are realized, under the constraining influence of a macro-semiotic system, the elements themselves will be subject, within each triad, to downward determinative relations. If we focus on the relations between elements within a triad, then we can see that Signs, Objects, and Interpretants constrain each 97
There may, however, be another sort of ‘vertical’ element of irreducibility, not so much of the Sign triad in itself but in relation to its action in processes where feelings and consciousness of the organism are in focus (cf. Emmeche 2003), but this is not our concern here.
Chapter 7. Emergence of semiosis: a general model
177
other’s behaviors. It is due to determinative relations from chains of triads to triads, and from triads to the elements themselves, that a Sign can act as a medium for communicating the form of an Object to an Interpretant. It seems, thus, that an interpretation of DD in terms of formal constraints applies smoothly to the determinative relations in triadicdependent processes. Surely, a proper interpretation of DD in semiotic phenomena demands more elaboration. But we shall leave this issue for future works.
Is the structure of semiotic systems or processes unpredictable? As to our eighth question, the actualized processual structure of triads and chains of triads in a particular complex evolutionary or developmental system can be indeed regarded as unpredictable.98 This is consistent with the fact that Peirce advocated that indeterminism, spontaneity, and absolute chance are fundamental factors in the universe. Thus, the behavior of the elements in a semiotic process may also be unpredictable from the behaviors they may exhibit in simpler systems. In a Peircean framework, we can claim, thus, that semiosis is a process the structure of which is in principle unpredictable due to the indeterministic nature of the evolutionary process. This argument is grounded on the Peircean thesis of tychism, the metaphysical defense of “absolute chance” as a real factor in the universe (see Murphey 1993; Potter 1997). Tychism plays an essential role in Peirce’s account of cosmological evolution, to the extent that he regarded it as the only explanation of the multiplicity and irregularity found in the universe. 98
This does not preclude, of course, that within domains of science, for particular systems involving semiotic processes, predictions can be made on normal pragmatic grounds, as one of the major aims of physical and biological sciences. Such fallible predictions are often practically achievable, testable, and limited by the measurement errors on initial conditions, etc. Similarly, you cannot predict which mutations may appear in a biological population subjected to a selective pressure to allow for specific adaptations, but you may still predict aspects of such processes given specific information about the system and its conditions. What we discuss here is the possibility of predicting all features about the emergence of a semiotic process from its simpler components.
178
Genes, Information, and Semiosis
The most important point for our arguments here is that, according to a Peircean evolutionary cosmology, everything should be explained as a product of an evolutionary process which has states of indetermination and chance as its starting points. In a paper about tychism and mental processes, Pape (2002: 226) comments that “matter, time, space, and the laws of nature themselves — they all have to be explained as emergent regularities of interaction arising from a state of indeterminateness”. This suggests, once again, the compatibility of emergentist thought with central doctrines in Peirce’s metaphysics, as synechism and tychism. Consider, moreover, that Peirce’s categories constitute a system of necessary presupposition (see Hausman 1993: 97), and, thus, it is impossible to conceive thirdness without secondness, and secondness without firstness. Therefore, as firstness entails indetermination, novelty, independence, and, consequently, unpredictability, the latter becomes a necessary component also in thirdness, and, thus, in semiosis.
3. Concluding remarks We drew on Salthe’s hierarchical structuralism to propose a model for explaining the emergence of semiosis in semiotic systems. According to this model, semiosis is conceived as a systemic process at a focal level, in which chains of triads are instantiated as a result of the interaction between potentialities established at a micro-semiotic level (initiating conditions) — containing potential Signs, Objects, and Interpretants — and the regulatory, selective influence of a macrosemiotic level (boundary conditions) — corresponding to networks of chains of triads. We intend that the questions we raised in order to characterize semiosis as an emergent process in a precise manner, the answers we proposed, and the modeling of this process based on Salthe’s hierarchical structuralism, contribute to a consistent treatment of the emergence of semiosis in the context of Peirce’s metaphysics and evolutionary cosmology. An understanding of how semiosis emerges in evolution has to offer, in our view, proper answers to the set of questions we systematized in this chapter, and, furthermore, has to work with adequate tools for modeling such an emergent event, such as those advanced by Salthe in his basic triadic system.
Chapter 7. Emergence of semiosis: a general model
179
The arguments developed above also lead to the conclusion that a strong emergence account can be supported in the case of semiotic phenomena. In conformity to Peirce’s theory of Signs, this theory should include (1) a concept of irreducibility based on the nondeducibility of the behavior of Signs, Objects, and Interpretants in semiotic relations from their possible behaviors in simpler relations, and (2) a concept of in principle theoretical unpredictability of the structure of semiotic processes, based on the doctrine of tychism. It is in terms of this account of emergence that we claim that the model developed in this chapter can explain the emergence of semiosis, as the instantiation of chains of triads at a focal level (defined by a particular research target). Semiosis is, in these terms, the product of an interaction between processes taking place at a micro-semiotic level, which establishes initiating conditions for semiosis, and a macro-semiotic level, a network of chains of triads which embed the semiotic processes at the focal level, and establishes boundary conditions for their actualization.
Chapter 8 Levels of semiosis in the genetic information system
1. The micro-semiotic level: strings of DNA as potential Signs A set of three nucleotides (codon) in an open reading frame (ORF) of a coding string of DNA is treated, in our analysis, as a potential Simple Sign, i.e., as a Simple Sign which is not involved at a particular time t in an effective triadic process involving Objects and Interpretants, which is not partaking in effective semiosis, but potentially can do so. In a similar way, we will refer here to a potential Composite Sign.99 Let us consider the string of DNA corresponding to the fibronectin (FN) gene (with all its exons and introns) in a given cell (Figure 4). In this case, each codon in the FN gene is a potential Simple Sign standing for one specific amino acid as its potential Simple Immediate Object, given the rules of the genetic code. The FN gene is a potential Composite Sign in DNA, which can have a range of effects on cells as interpreters. In a given cell, one of these effects will be actualized, i.e., a Dynamical Interpretant of a Dynamical Object, a FN isoform encoded by the FN gene. When the FN gene, as a potential Composite Sign which can undergo alternative splicing, is actualized, a particular splicing pattern will be selected among all possible patterns that might be selected. Thus, a particular sequence of amino acids (Composite Immediate Object) will be selected and reconstructed (Dynamical Interpretant) among all possible sequences of amino acids that might be synthesized in that cell type and developmental state (the range of interpretability, the Immediate Interpretant of the Composite Sign). 99
It is in this sense that we will talk about ‘potential genes’ in the following sections. Cf. Jablonka’s (2002: 586–587) hypothetical example of proto-cells in which DNA was not used as an ‘informational resource’ but as a high-energy storage polymer. If, in such proto-cells, a given stretch of a storage DNA polymer had, by chance, the precise sequence coding for, say, fibronectin, this would have ‘... of course [...] no special consequences for the proto-cell, since there is no cellular system that can interpret this sequence in a specific way’ (id. ibid.). That is, that coding stretch of DNA would always remain a potential Composite Sign, a potential gene, and, since it would never get actualized, it would mean nothing to a cell, it would be no effective information at all, precisely for the lack of a Dynamical Interpretant.
184
Genes, Information, and Semiosis
That Immediate Object will, then, indicate a particular Dynamical Object, say, the fibroblast FN isoform. To understand the idea of ‘potentiality’ in this explanation, consider, for the sake of our argument, a cell in a given state at time t in which the FN gene is not being transcribed, i.e., the codons in the FN gene are merely potential Simple Signs, and the FN gene as a whole, a potential Composite Sign. This is a situation in which we can talk only about a micro-semiotic level as a set of initiating conditions for effective semiotic processes which are not instantiated at that time t. After all, in a string of DNA which is not transcribed in mRNA , or in a string of mRNA that is not translated into a polypeptide, codons are not effectively acting as Signs. In these circumstances, codons can potentially be Signs of an Object for an Interpretant. Free amino acids, by their turn, can potentially be Objects of that Sign for an Interpretant. Finally, the potential Interpretant amounts to the potentiality of a specific sequence of amino acids (Composite Immediate Object) being reconstructed from a Composite Sign in DNA, by means of the processes of transcription, RNA processing, translation etc. To inquire further into the idea that a non-transcribed reading frame in DNA is nothing but a potential Composite Sign, consider a hypothetical situation in which the FN gene remain non-transcribed in all states of a given cell, at any given time t. In this case, the FN gene and the codons composing it will never effectively act as Signs; rather, they will remain potential. The string of DNA containing these codons will always remain as a silent structure that might — potentially — engage in the process of becoming effective information. As we interpret what is a gene from a semiotic standpoint, an exciting conclusion is suggested from the very beginning, namely, that information in a gene is not an entity, but a process. If this is the case, then DNA will not ‘harbor’ or ‘carry’ information as sequences of nucleotides (which are patterns composed of entities, after all), but only the potentiality of engaging in processes by means of which the form of an Object can be communicated to an Interpretant, i.e., what we call here ‘potential information’. We are moving towards a reinterpretation of what is information in a cell that hopefully avoids a problem detected by Oyama (2000), namely, that genetic determinism is implied by the way we represent genes as they carried information for the development (or functioning) of an organism. We will come back to this point later.
Chapter 8. Levels of semiosis in the genetic information system
185
The idea that a string of DNA encoding the sequence of amino acids of a protein is just a potential Sign when it is not being transcribed can benefit from the notion of experience as “... traces of particular significant interactions between a system and its surroundings that for some period are represented within the system” (Emmeche 2003: 325). Experiences are, thus, “... fossilized signs [...] or quasi-stable forms of movement that organize the system’s past forms of movement in such a way as to have significant consequences for the system’s future movement” (Emmeche 2003: 327). Similarly, DNA sequences can be regarded as ‘fossilized’ Signs that represent within the system past interactions with its surroundings in such a way as to have significant, i.e., adaptive consequences for the system’s future dynamics. Thus, in biological systems like the cell, experiences are, among other things, the genetic ‘fossils’ in DNA witnessing the specific proteins that were functionally participating in earlier ancestor cell lines to maintain the metabolic form of movement. (Emmeche 2003: 328)
DNA sequences are just physical carriers of past experience, i.e., potential Signs. When they are put into effective action in a cell (rather than act on their own), they become part of an effective triadic process, through which they can have an effect on a cell by irreducibly involving also Objects and Interpretants. If we go on with the analogy, we will be able to see that an unexpressed gene in a cell is a potential Sign as much as an undiscovered fossil deep down in a mountain. We can postulate that a hypothetical fossil buried in a rock but never seen and interpreted before is a Sign on the grounds of our previous experiences of the Sign action of fossils: we have already a habit of interpreting patterns of rock as Signs of an ancient fish or dinosaur. Similarly, we have enough knowledge of genes to postulate that particular tokens of genes may be potential Signs, i.e., they may be interpretable by the cell as Signs for the process of synthesis of a specific protein, in response to some necessity. Pragmatically, also, a potential Sign is known by its effects, these being as hypothetical as the very Sign; yet we judge these future effects as real based upon an inference that relate past tokens of similar types of Signs to their Objects and Interpretants. The unseen gene is as silent as the unseen fossil. Collateral evidence — about expressed genes or neighboring disclosed
186
Genes, Information, and Semiosis
fossils — supports our claim about the possible existence of unseen, silent, potential Signs. A potential Sign is information that does not — yet — have an effect on the interpreter, but has the power to do so in the future. In living systems, experience became so intensified in semiotic terms that it can reach forward in time. This is true not only of experience in quite complex but also in much simpler living systems, where experience can take, for instance, the form of a genetic memory (compare the term ‘form’, as defined in Ch. 4), which, given the stability of DNA, can represent traces of significant types of interactions between a living being and its surroundings for quite long periods, reaching the future not only in the restricted time scale of somatic life, but also in the much more expanded time scale of evolutionary processes. The representation of experience as a quasi-stable dynamical pattern (Emmeche 2003) renders the system anticipatory, endowing it with the capacity of operating with models of possible future states. It is not the case at all that we are claiming that the genetic information system might be prescient in some sense or another. It is just that, if the selective regimen for a given lineage remains stable in the relevant variables, past selective events — i.e., past events of differential survival and reproduction — endow the future generation with potential Signs in DNA that are traces of adaptive interactions between a system and its surroundings, and are likely (but not surely) to create conditions for successful future interactions (see Ch. 6).
2. Transcription, RNA processing, and protein synthesis as processes of gene actualization Transcription, RNA processing, and protein synthesis can be understood, in semiotic terms, as processes of actualization of potential signs in protein-coding genes. Consider, for instance, a given hepatocyte h, in which the FN gene is transcribed and the corresponding mRNA, after cell type-specific splicing, is translated into the hepatocyte-specific FN isoform. These processes actualize potential Signs in a string of DNA, turning them into actual Sign processes, Signs in effective action in an organism. When put into action, the nucleotide sequences in that string of DNA become part of effective
Chapter 8. Levels of semiosis in the genetic information system
187
semiosis, a triadic-dependent process by means of which the FN gene as a Composite Sign indicates, through a process involving the actualization of each Simple Sign composing it, the functional hepatocyte FN isoform as a Dynamical Object. This FN isoform has in turn an effect on the organism in which it is expressed (its Dynamical Interpretant), participating in its adaptive interactions with its surroundings, and, thus, contributing to the presence of the FN potential gene in the next generation in a high frequency. The actualization of a potential gene in a string of DNA depends on boundary conditions established by a higher-level semiotic network, a network of signalling processes, involving many chains of triads, which will regulate or coordinate gene expression, ultimately determining the likelihood of transcription of a given gene, or splicing of a given pre-mRNA according to a particular pattern, or chemical modification of a given protein in a manner that modulates its function in a particular way (e. g., a phosphorylation), and so on. A variety of regulatory mechanisms studied in cellular and molecular biology can be thus interpreted as composing a macro-semiotic environment establishing boundary conditions which will downwardly determine which potential genes in a string of DNA will be turned into actual genes, into genes in effective action in a cell. These mechanisms determine which sequence of amino acids will be actually reconstructed (Dynamical Interpretant) among all those that might be reconstructed (Immediate Interpretant, the range of interpretability of a Sign) out of a string of DNA (Composite Sign). This shows how several complexities involved in the gene concept and gene expression can be introduced in our analysis: boundary conditions established by this macro-semiotic environment will determine, for instance, which stretch of DNA will be read (e. g., allowing for an analysis of transcription of overlapped or nested genes), which pattern of RNA splicing or RNA editing will be instantiated in order to produce a particular mature mRNA (allowing for the subtleties of alternative RNA splicing or RNA editing to be taken into account), which functional protein will be effectively constructed by the cell (allowing for chemical and/or structural modifications suffered by the primary amino acid sequence of a protein to be considered), and so on. However, we will avoid introducing a great deal of complex details now; rather, we will concentrate on an analysis of transcription and translation, since our
188
Genes, Information, and Semiosis
goal here is to establish a set of concepts, tools, and procedures for the analysis of information systems in living beings, not to provide an exhaustive analysis of the host of processes involved in these systems, not even at the cellular level.
3. Semiotic analysis of transcription The first step in the actualization of potential Signs in a string of DNA is transcription. This process turns potential Signs in DNA into potential Signs in pre-mRNA. It is easy to see that Signs in premRNA are still potential, since they will become part of actual triads only if they are effectively translated.100 It can be the case, for instance, that a given codon in pre-mRNA is located in an alternatively spliced exon that is eliminated from the final transcript in a given cell type, developmental stage or age. In this case, the actualization process is not completed and that codon remains in the condition of a potential Simple Sign. Consider, for instance, Exons EIIIA and EIIIB in the FN gene. Since they are spliced out of FN mRNA in hepatocytes, the codons in those exons are never actualized, remaining as potential Signs in this cell type. In fibroblasts, however, these potential Signs will be indeed actualized. Transcriptional control is the major mechanism for regulating the production of a protein encoded by a given stretch of DNA, involving both repression and activation of specific genes in response to signals originating from the cell itself and, more often, from the extracellular environment. In terms of the general model presented above, this means that, as a first step in the actualization of potential Signs at the micro-semiotic level, transcription is constrained by boundary conditions established by networks of chains of triads (macro-semiotic level) which ultimately determines the likelihood of transcription of a given potential Sign in DNA. Transcriptional regulation amounts to the selection (by the macro-semiotic environment) of a specific chain of triads to be actualized, among many potential chains that might be actualized at a given moment. Furthermore, transcriptional regulation 100
For the sake of clarity, it is important to emphasize that this claim applies to the event in which nucleotide sequences in mRNA become part of actual triads qua Signs, since, if they were transcribed, they were already part of actual triads, but qua Immediate Objects.
Chapter 8. Levels of semiosis in the genetic information system
189
is not at all a case of random selection, but rather the result of mechanisms selected in the course of the evolution of a lineage, due to the differential fitness of varying responses of the cellular regulatory systems (as a cellular macro-semiotic environment) to boundary conditions or selective regimens established by the environment outside the cell, and outside the organisms as a whole. Let us now analyze in more detail transcription as a semiotic process. We will consider here two views of the processes at stake, the ‘horizontal’ and the ‘vertical’ perspectives. If we take a ‘horizontal’ view of transcription, we will see a mechanistic process in which RNA polymerase moves along a string of potential Simple Signs in DNA, triggering subsequent semiotic events, in which those potential Signs become part of triads including Objects and Interpretants. Let us focus first on a Simple Sign in DNA, i.e., a set of three nucleotides in a coding region (Figure 12). The Simple Immediate Object, by its turn, is a set of three nucleotides in mRNA. In our example, a codon in the FN gene is a potential Simple Sign that is actualized when that gene is transcribed.101 As we argued above, the Immediate Interpretant of a Simple Sign is the range of interpretability established by the rules of base pairing, and its Dynamical Interpretant is the realization of a particular rule by means of which specific nucleotides in DNA determine specific nucleotides in mRNA. When a triad in transcription is actualized, the interpretative subsystem of the cell (as a global interpreter), RNA polymerase, moves to the next codon in the string of DNA, i.e., to the next Simple Sign. If we take a ‘vertical’ view, we will consider the relationship between semiotic processes in transcription and translation. Then, we will see a dynamical process in which the Simple Immediate Object of each triad actualized in each step of transcription, i.e., a three-nucleotide sequence in mRNA, becomes a potential Simple Sign in the next Sign process in the actualization of a gene, translation (Figure 13).102 101
But notice that, even though an actualization of potential triads indeed takes place in transcription, this is just a first step in the process of actualization of a potential gene in DNA, which will involve several other steps. The actualization of a potential Sign in DNA results in another potential Sign, now in pre-mRNA. 102 Just for the sake of the argument, we are skipping RNA splicing here. It is as if we were analyzing a string of DNA resulting in a pre-mRNA with only one mature mRNA as the result of its processing. This doesn’t mean, however, that we don’t take RNA splicing into account in our analysis, as the previous comments about the FN gene in this very section show. If we consider that splicing patterns
190
Genes, Information, and Semiosis
Figure 12. A ‘horizontal’ view of transcription. The letter ‘N’ is used along this book to identify the elements of a triad at the level of transcription. NS: Simple Signs in DNA (codons); NO: Simple Immediate Objects in pre-mRNA (codons); N I: Immediate Interpretants in transcription, the range of interpretability established by the rules of base pairing by which specific nucleotides in DNA determine specific nucleotides in mRNA. RNA polymerase is the interpretative system performing the transcription. The arrows represent the movement of the interpretative subsystem, RNA polymerase, to the next Simple Sign, when a triad is actualized in transcription.
4. Semiotic analysis of protein synthesis We can now go on to analyze protein synthesis in semiotic terms, considering both the recognition of codons in mRNA by particular tRNAs and the attachment of appropriate amino acids to specific tRNAs. Let us consider, first, the attachment of amino acids to tRNAs. In this case, the Sign in a given triad is the three-dimensional structure of a particular tRNA, which is recognized by the interpretative system in this process, a specific aminoacyl-tRNA synthetase. The Simple Immediate Object in each triad is a specific amino acid, which is also recognized by aminoacyl-tRNA synthetase on the grounds of its threedimensional structure. The aminoacyl-tRNA synthetase establishes a relationship between one amino acid (the Simple Immediate Object) and all its cognate tRNAs (with specific three-dimensional structures as Signs) due to its capacity of specific recognition. This enzyme actualizes, thus, one of the rules expressed in the genetic code, the are cell type-, development-, and age-regulated by mechanisms involving signalling pathways, we will see that semiotic processes play a role also in premRNA splicing. For reasons of space, we will not explore this avenue in the scope of this book, leaving it for subsequent works.
Chapter 8. Levels of semiosis in the genetic information system
191
Immediate Interpretant of a Simple Sign. This actualization is the Dynamical Interpretant.
Figure 13. A ‘vertical’ view of the relationship between Sign processes in transcription and translation. The letter ‘T’ is used along this book to identify the elements of a triad at the level of translation. The letter ‘N’ and the symbols NS, N O, and NI are used as indicated in the caption of Figure 12. TS: Simple Signs in mRNA (codons); TO: Simple Immediate Objects (particular tRNAs with specific anticodons); TI: Immediate Interpretants in translation, the range of interpretability established by the rules of base pairing by which specific nucleotides in mRNA are paired with specific nucleotides in tRNA. For more details on the semiotic analysis of translation, see next section. The arrow indicates that three-nucleotide sequences in mRNA, the Simple Immediate Objects in transcription, become potential Simple Signs in translation.
With regard to the recognition of codons in mRNA by particular tRNAs, the Simple Signs are three-nucleotides sequences in mRNA (codons), the Simple Immediate Objects are particular tRNAs with specific anticodons, and the Immediate Interpretant of the Simple Sign is the range of interpretability established by the rules of base pairing by which specific nucleotides in mRNA are paired with specific nucleotides in tRNA, with the caveat that nonstandard base pairing often occurs between codons and anticodons. The Dynamical Interpretant is the actualization of a specific base pairing. When a triad in this step of translation is actualized, the ribosome, as an interpreta-
192
Genes, Information, and Semiosis
tive subsystem, moves to the next codon in the string of mRNA, i.e., to the next Simple Sign. Since several processes in protein synthesis, including translation, protein folding, association of different polypeptide chains, and posttranslational chemical modifications are often regulated, this step in the actualization of potential Signs in DNA is also constrained by boundary conditions established by networks of chains of triads at a macro-semiotic level, which select determinative relations between S, O, and I at the micro-semiotic level. Again, we can see these processes in a ‘horizontal’ or a ‘vertical’ view. Consider, first, the semiotic processes involved in the recognition of codons in mRNA by tRNAs. If we take a ‘horizontal’ view of this process, we will see, as in transcription, a mechanistic process consisting in the triggering of a sequence of Sign processes by the movement of ribosomes along strands of mRNA (Figure 14).
Figure 14. A ‘horizontal’ view of translation events in the ribosome. The letter ‘T’ and the symbols TS, TO, and TI are used as explained in the caption of Figure 13. The ribosome is the interpretative system performing this step in protein synthesis. The arrows represent the movement of the ribosome to the next Simple Sign, when a triad is actualized in translation.
If we take a ‘vertical’ view of the relationship between semiotic processes in translation and in aminoacyl-tRNA synthesis, we will see a process in which the Simple Immediate Object of a triad, a tRNA with an anticodon that matches a codon in mRNAs, is also a potential
Chapter 8. Levels of semiosis in the genetic information system
193
Simple Sign103 in the semiotic process in which a specific aminoacyltRNA is synthesized (Figure 15). Other steps in protein synthesis can also be analyzed semiotically. For instance, protein folding, at least when it involves molecular chaperones, a special class of proteins that help guide the folding of many proteins, is regulated by processes involving signalling pathways. Nevertheless, we will not develop a semiotic analysis of this process in the context of this work.
Figure 15. A ‘vertical’ view of the relationship between semiotic processes in translation and in aminoacyl-tRNA synthesis. The letter ‘T’ and the symbols TS, T O, and TI are used as explained in the caption of Figure 13. The letter ‘H’ is used to identify the elements of a triad at the level of aminoacyl-tRNA synthesis. HS: Signs: three-dimensional structure of tRNAs; HO: Simple Immediate Objects: specific amino acids; HI: Immediate Interpretant, the range of interpretability of each codon as a Simple Sign, established by the rules of the genetic code. The arrow indicates that the same element, a tRNA with a specific anticodon that matches a codon in mRNAs, plays the different functional roles of Simple Immediate Object and Simple Sign in different triads.
Finally, it is worth discussing start and stop codons in the context of the semiotic analysis developed here. Translation is always initiated by the recognition of a start codon in mRNA, usually AUG, by a 103
We should still talk about potential Signs in this case because the process of actualization of a potential gene can still be interrupted, since it depends, for instance, on the availability of specific amino acids. Consider, for instance, the case of a starving animal which can lack some amino acids necessary for protein synthesis.
194
Genes, Information, and Semiosis
tRNA carrying the amino acid methionine. Translation is, therefore, a semiotic process with a peculiar characteristic: it typically begins with the same Simple Sign (AUG) and always with the same Simple Immediate Object (methionine). This Immediate Object, however, is in most cases subsequently eliminated from the sequence of amino acids which indicates the Dynamical Object of the Composite Sign, and, therefore, we have here an Immediate Object which is not really related to the semiotic availability of the functional protein indicated by a gene. In this case, the Dynamical Interpretant is the actualization of a rule of the genetic code, by which AUG usually encodes methionine, and the Dynamical Object is the instruction that translation should be initiated. Stop codons (UAA, UAG, and UGA), in turn, are usually involved in the termination of the semiotic process of translation. None of the stop codons is recognized by a tRNA. Rather, they are recognized by proteins called ‘release factors’, which act at the ribosomal site occupied, in the case of other codons, by an aminoacyl-tRNA. When a release factor binds this site, a molecule of water, instead of an amino acid, is added to the growing polypeptide chain, resulting in its release from the last tRNA. In this case, the Dynamical Object of the semiotic process is the instruction that the process should be interrupted, and this Dynamical Object is made semiotically available by the fact that a molecule of water, rather than an amino acid, is the Immediate Object of the Simple Sign at stake.
5. A global picture It is time, then, to look at the processes discussed above from a global perspective, which will allow us to use the detailed analysis we carried out to reach, as an overall conclusion, the initial semiotic analysis presented in Ch. 6. Figure 16 shows a complete view of all the semiotic processes involved in the actualization of potential Signs in DNA that were explained in the previous sections. It is important to notice that the chains of triads are, as shown in the figure, arranged vertically, going from transcription to translation and aminoacyltRNA synthesis, each codon a time. The actualization of potential Signs in DNA is, indeed, a complex process including an impressive array of subsidiary semiotic processes, described in a stepwise manner
Chapter 8. Levels of semiosis in the genetic information system
195
above and expressed as a whole in Figure 16. This is in accordance with the general idea that in living nature there are different levels of handling of information, i.e., generation, translation, coding, recoding, and interpretation of Signs. From the view of the semiotic processes involved in the actualization of potential Signs in DNA shown in Figure 16, we can obtain a global picture (Figure 17) that corresponds to the semiotic analysis presented in Ch. 6. As we argued above, if genes are treated as Signs, they can only have an effect on a cell through a triadic process which is genetic information, and involves an irreducible relationship between three elements: the Composite Sign, which is a string of DNA, and can be transcribed into RNA, processed, and, in the case of protein-coding genes, translated into a protein (or a polypeptide), by means of the semiotic processes analyzed above and shown in Figure 16; the Composite Immediate Object, which is, in the case of proteincoding genes, a linear sequence of amino acids, and, in the case of RNA genes, a linear sequence of ribonucleotides; and the Immediate Interpretant of a Composite Sign, which is its range of interpretability, i.e., the possibilities of reconstruction of sequences of amino acids or sequences of RNA (Immediate Objects) from that Sign in DNA. The Dynamical Interpretant of a Composite Sign corresponds, in turn, to the effective reconstruction of a sequence of amino acids or a sequence of ribonucleotides from a Sign in DNA.104 As the global picture in Figure 17 illustrates, a model of genetic information interpreted as a Sign process can be obtained by the summation of lower-level semiotic processes, involving triads of which the codons in DNA (Simple Signs) are the first correlates. In this sense, the Composite Sign (a stretch of DNA) and the Composite Immediate Object (a linear sequence of amino acids or ribonucleotides) can be treated as resulting from an accumulative process of interpreting Simple Signs (codons in DNA) of Simple Immediate Objects (amino acids or ribonucleotides).
104
This analysis should be made somewhat more complex to accommodate mRNA editing, a process in which individual bases are added to or deleted from mRNA during processing (see Chapter 2). In this case, Simple and Composite Signs in mRNA are changed in such a manner that the Composite Immediate Object has a different sequence, in the end, from that which is semiotically available in DNA.
196
Genes, Information, and Semiosis
Figure 16. Whole view of the semiotic processes involved in the actualization of potential Signs in DNA. Letters ‘N’, ‘T’, ‘H’, and symbols NS, NO, NI, TS, TO, TI, H S, HO, HI are used as explained in Figures 16, 19, 21. Dashed arrows represent relationships between transcription and translation, and translation and aminoacyltRNA synthesis, as explained in Figures 19 and 21. Continuous lines indicate the horizontal and vertical views explained in the text. Asterisks indicate signalling processes that can be analyzed semiotically, but were not addressed here for reasons of space.
Chapter 8. Levels of semiosis in the genetic information system
197
Figure 17. A global picture of genetic information as a triadic process. The letters ‘N’ and ‘H’ indicate transcription and aminoacyl-tRNA synthesis, respectively. The symbols NS and HO indicate a Simple Sign in DNA (a codon) and a Simple Immediate Object in the synthesis of aminoacyl-tRNA, that is, a specific amino acid. By means of the specificity of recognition of amino acids and tRNAs by aminoacyltRNA synthetases, the rules connecting 61 codons and 20 amino acids in the system of constraints expressed in the genetic code (the Immediate Interpretant of the Simple Sign) obtain. In the picture, it is indicated by HI. By the summation of individual chains of triads, we obtain a global picture of genetic information as a semiotic process involving a gene as a Composite Sign, COMPS, i.e., a string of DNA; COMP O, the Composite Immediate Object of the gene, i.e., the linear sequence of amino acids in a protein (or polypetide), in the case of a protein-coding gene; and the Immediate Interpretant of the Composite Sign (COMPI), i.e., its range of interpretability, the possibilities of reconstruction of sequences of amino acids from that Sign in DNA. The expression ‘COMP’ stands here for ‘Composite’. This global model is equivalent to the picture shown in Figure 6.
The Dynamical Object of a Composite Sign in DNA is a functional, folded, and chemically modified protein, which can exert a particular effect (the Dynamical Interpretant of the Dynamical Object) on a cell or organism of which the cell is part, among a range of possible effects (the Immediate Interpretant of the Dynamical Object). It is only then that a potential Sign, a potential gene in DNA, turns into an actual Sign, a gene effectively involved in the Sign process we interpret here as genetic information. To put it differently, the full actualization of a string of DNA, which is only a potential Sign, demands the ultimate indication of a Dynamical Object, a functional protein, by the Composite Immediate Object, a polypeptide chain (in the case of a protein-coding gene). Only then the path from potential to effective information is completed in the genetic system. The actualization of potential Signs in
198
Genes, Information, and Semiosis
DNA requires a series of interpretative subsystems, such as RNA polymerases, ribosomes, aminoacyl-tRNA synthetases, etc. The regulatory influence of the macro-semiotic level, as a network of signalling processes, on interpretative subsystems, and, thus, on transcription, splicing, translation, shows that, in the end, we have to consider the whole cell as ultimately participating in the network necessary for the actualization of potential genes in DNA (see Ch. 6). The cellular network of chains of triads is, in turn, highly responsive to environmental factors, given the semi-open nature of living systems. Finally, we should consider how the process of actualization of a potential gene in DNA can be embedded into the model of semiosis in three levels shown in Figure 10.105 As Figure 18 shows, potential Signs, potential genes in DNA, are actualized in response to regulatory dispositions arising from a network of signalling pathways that elicit cellular specific responses to Signs arising from the extra- or intracellular environment.106 A controlled, regulated answer by a cell is impossible without signalling. When a particular gene product is needed, a signal from the environment activates the expression of a given gene by means of signalling mechanisms. The cell, as an interpreter, answers to an environmental cue by means of a specific alteration of its internal states, triggered by a whole network of signal transduction culminating in a change at some level of gene regulation (For a biosemiotic analysis of signal transduction systems, see Bruni 2003, 2007; Queiroz, El-Hani 2006b; El-Hani, Arnellos, Queiroz 2007). These relations cannot be understood only in terms of molecular interactions taking place in the network of signal transduction, because the latter crucially involves semiotic processes, as the widespread usage of information talk in modeling and explaining signalling pathways clearly suggests. Through signalling pathways, cells
105
In the following arguments, we will not focus on genes which are constitutively expressed, such as housekeeping genes, but rather on genes that can be turned on or off depending on the context in which a cell is embedded. 106 We use the expression ‘extra- or intracellular environment’ mostly for the sake of simplicity. There is, in fact, a hierarchy of ‘contexts’, ‘environments’, or, in our own terms, semiotic levels that can direct gene expression (i.e., establish boundary conditions for the selection of potential genes in DNA), ranging from systems of gene-gene interactions to organisms, and passing through nucleus, cytoplasm, cell, cell surface, extracellular matrix, morphogenetic fields, collective condensations of cells (blastemas), organs, etc. (see, e. g., Hall, 2001).
Chapter 8. Levels of semiosis in the genetic information system
199
Figure 18. The process of actualization of a potential gene in DNA embedded into the model of semiosis in three levels shown in Figure 10. COMPS, COMPO, COMPI, N S, NO, NI, TS, TO, TI, HS, HO, and HI are used as explained in the previous figures. COMP S1-n stands for potential Signs at the micro-semiotic level. ∆t1 and ∆t2 in the upper part of the figure indicates the diachronic nature of the semiotic processes involved. At the focal-level of semiosis, we show a chain of triads at the focal level and the global picture shown in Figure 17.
200
Genes, Information, and Semiosis
are able to interpret Signs from the extra- and intracellular environment as meaning something beyond the chemical carrier of the Sign itself. Thus, the presence of an antigen bound to a membrane receptor may mean, for instance, that the organism is under the threat of a pathogen. In response to such an interpretation of the environment, a signal transduction cascade can be activated in B-cells, for instance, leading to the actualization of genes associated with B-cell activation, and, thus, to specific effects on the B-cell, making it, say, engage in a process of presenting peptides derived from the antigen in its cell surface, where they can be recognized by T-cells, leading to helper T-cell activation (which in turn leads to full B-cell activation). As molecules come to mean something else than just being molecules (in our example, the threat by a pathogen) and cells use them as parts of a process of interpreting its circumstances, something more than chemistry is going on here. Needless to say, there is nothing supernatural going on; it is just the case that information plays a fundamental role in the lives of organisms, and information can be interpreted as a triadic-dependent process, as we have argued throughout this book. Signalling pathways in a cellular system play the role of establishing boundary conditions to processes at the focal- and microsemiotic levels, downwardly selecting particular strings of DNA, potential genes, to be actualized, among all the potential Signs at the micro-semiotic level that might be actualized at a given time t. It is the actualization of a specific set of potential genes (which can — but not necessarily should — include only one element, as shown in the hypothetical case in Figure 18) that allows the cell to answer to a given signal in a specific way, by means of Dynamical Objects and their Dynamical Interpretants. We argued above that our analysis leads to a model in which genetic information is explained as a Sign process obtained by the emergence of chains of triads at a focal-semiotic level. We should remember, now, that we are using the term ‘emergence’ here in a technical sense, as explained in detail in Ch. 7. In this sense, genetic information, as an emergent semiotic process, is interpreted as a higher-level process related in a certain way to the microstructure of a semiotic system (as defined in Ch. 7). In that Chapter, we developed a strong account of the emergence of semiotic phenomena, in which semiosis is treated as being irreducible due to the non-deducibility of the behavior of Signs, Objects, and Interpretants in semiotic relations from their possible behaviors in simpler relations, and in principle
Chapter 8. Levels of semiosis in the genetic information system
201
theoretically unpredictable, in virtue of the unpredictability of its structure, which follows from Peirce’s doctrine of tychism. For our purposes here, we can leave this sort of unpredictability aside, since emergent processes can be, at the same time, theoretically unpredictable and inductively predictable (see Ch. 7, note 88), after they emerged for the first time, as a result of synchronic determination: if emergent processes are synchronically determined by the microstructure of a system, we can inductively predict that they will emerge when the right structure is present. But this doesn’t mean that they are predictable tout court, since the very structure by which they are determined can be theoretically unpredictable, in the case of semiotic processes. As we are dealing here with a well-defined class of semiotic systems, the structure of which is well-known, the emergence of genetic information can be inductively predicted from the instantiation of a give structure in a living system, as a semiotic system (while it remains theoretically unpredictable).107 Therefore, we will focus our arguments on irreducibility here. In the model developed in Ch. 7, the emergence of semiosis is explained as the instantiation of chains of triads at a focal level, and this instantiation is, in turn, treated as the product of an interaction between processes taking place at a micro-semiotic level, which establishes initiating conditions for semiosis, and a macro-semiotic level, a network of chains of triads which embed the semiotic processes at the focal level, and establishes boundary conditions for their actualization. It is clear, then, that we cannot deduce that a given pattern of genetic information, of semiotic processes by means of which a set of potential genes will be actualized, from knowledge about genes only. We need to consider the boundary conditions established by the network of signalling processes in a cell, as a macrosemiotic system which responds, in turn, to environmental cues. In the whole process of gene actualization, it is the cell and, as Hall (2001) emphasizes, its immediately adjacent peri- and extracellular matrices that carry out the responses to environmental changes. That is, 107
But notice that, to predict the instantiation of a given structure in a semiotic system, one needs to consider it as a semi-open system in a given environmental circumstance. We cannot, for instance, predict the structure underlying a given pattern of gene expression simply from knowledge about the genome. We need to consider the environmental and epigenetic influences over the dynamics of the genome to make such an inductive prediction.
202
Genes, Information, and Semiosis
although potential genes in DNA are being actualized in the process of responding to a given intra- or extracellular signal, it is not the case at all that genes are in the command; rather, they are commanded by the interpretive mechanisms of the cell to supply the materials needed for such a response. This is yet another sense in which genetic information, as a semiotic process, is irreducible to sequences of nucleotides in DNA. The behavior of such sequences, as Signs, is not deducible from the behaviors they might show in simpler, nonsemiotic systems, in which these sequences would not act as Signs for particular Objects, but would be merely molecules, as any other molecules. Furthermore, in non-semiotic systems, these Signs would not be integrated into triads, and, consequently, into chains of triads, and, thus, would not show the specific behavior it shows in a semiotic system, realizing genetic information as a semiotic process. And the same kind of analysis can be carried out for Objects and Interpretants. We should be careful, then, not to focus our understanding of the information networks and pathways in a cell excessively on genes. Rather, we should always take in due account how epigenetic controls are imposed on genes, i.e., how a network of interactions between genes, proteins, RNAs, and other molecules has a determinative influence on gene regulation and expression. This is the subject of epigenetics, which can be described as a new frontier in genetic research. As discussed in Ch. 2, we cannot simply equate biological function with protein coding by genes. Consider, for instance, that much of the complexity of several organisms, such as vertebrates, results mostly from changes in patterns of gene regulation, closely related to signalling pathways, which can be conceptualized as a macro-semiotic environment in which genes are embedded as one of the elements, and on which they crucially depend to turn from potentiality to actuality. Or, to put it differently, the composition, form, and function of the proteins which are expressed in a cell critically depend on cytoplasmic and extracellular signals which are interpreted by networks of proteins and RNAs in interaction both in the nucleus and in the cytoplasm. These networks can be even said to characterize the state of the cell as a whole, and, consequently, we can treat the macro-semiotic level as the very semiotic system in this case. This is consistent with Peirce’s idea that semiotic systems can be treated as the embodiment of semiotic processes (see Ch. 7).
Chapter 8. Levels of semiosis in the genetic information system
203
Obviously, the macro-semiotic network in a cell can and should be the object of semiotic analyses. Nevertheless, we shall leave this issue to future works. We will now turn to a discussion about what is information from the standpoint of the semiotic analysis presented in this book.
Chapter 9 Genes, information, and semiosis
1. What is genetic information? Let us now come back to the claims we put forward in the end of Ch. 6: Genes should be regarded as Signs in DNA, which can only have any effect on a cell through a triadic-dependent process (semiosis); this process is genetic information and involves more than just genes as Signs in DNA but also Objects and Interpretants; genetic information is the process by means of which a form in a Dynamical Object (a functional protein) is communicated to an Interpretant (the reconstruction of a specific sequence of amino acids in a cell) through Signs in DNA.
In this chapter, we will offer an extended interpretation of these claims as a result of the analysis of transcription and translation presented in the preceding chapters. According to the model developed in Ch. 7 and applied to the genetic information system in Ch. 8, the actualization of a potential gene is an emergent process at the focal level of semiosis, depending on two sets of constraints (see Figure 18). First, since a given organism, as a product of a historical process, does not and cannot contain each and every possible coding string of DNA but only a specific set of them, its response to a given extra- or intracellular signal is constrained by the very fact that it contains a restricted set of available potential genes in its genome. According to the model of the emergence of semiosis proposed here, this set of potential genes establishes initiating conditions for the organism’s response. Second, a network of signalling pathways at the macro-semiotic level also constrains the organism’s response, with the result that the response usually turns out to be quite specific to a given class of signals. In our model, this amounts to the establishment of boundary conditions by the macro-semiotic level, i.e., to a set of higher-level constraining conditions that result in the selective choice of a set (which can but not necessarily contain only one member) of potential genes available in that organism’s DNA to be actualized. To be sure, much work remains to be done in order to model semiotic processes at
208
Genes, Information, and Semiosis
signal transduction pathways, but we should leave a detailed treatment of the macro-semiotic level in cellular systems to future works.108 Here, we should now focus on answering what is, precisely speaking, information in this emergent process? Putting the concepts of information discussed in Ch. 4 to work, we can say that the actualization of a potential gene triggers a triadic-dependent process by means of which that gene has an effect on the cell. This process is effective information. A gene has an effect on the interpreter because it mediates, as a Sign, a process by which the form of a Dynamical Object (a functional protein) makes a difference to that interpreter. It is clear, then, that effective information — as defined here — is not contained in DNA109, but, rather, is a semiotic process which is irreducibly triadic, involving a gene in action, a dynamical Sign that has an effect on its interpreter by determining an Interpretant to stand in a similar relation to something else (the Object of the Sign) as that to which the Sign stands, thus mediating the relation between Object and Interpretant.110 In the context of our analysis, we can say that, when a cell, as an interpreter, responds to an environmental cue, by means of a set of signalling pathways that ends up altering its pattern of gene expression, triggering the actualization of a set of potential genes, what happens is that the interpretation systems of a cell are acting to create differences inside the cell in correspondence to differences in the external environment interpreted by it. In response to a state of the system plus its environment, a difference is established between two or more classes of gene expression patterns, in which different sets of potential genes are actualized, that is, become effective information, as 108
For a discussion about signal transduction pathways from the standpoint of biosemiotics, see Bruni (2003, 2007); El-Hani, Arnellos, Queiroz (2007). 109 As one of us wrote in a previous paper, “… biological information is not identical to genes or DNA […]. Information, whether biological or cultural, is not a part of the world of substance. It nevertheless depends on that world since it has to do with the pattern of substance, the way substance is organized or formed. […]. The dynamics of biological information belongs to another level of analysis than the dynamics of DNA-molecules. Biological information is expressed through signs and should be studied as such, i.e., as a special case of semiotics… ” (Hoffmeyer, Emmeche 2005: 29. Emphases in the original). 110 This shows how the metaphor that information ‘flows’ in a cell is inadequate, since it may lead to the misinterpretation that information is a sort of entity going from one place to another. See Hoffmeyer (2002).
Chapter 9. Genes, information, and semiosis
209
defined in this book. These differences in patterns of gene expression have an effect on the cell by altering its internal states, through the action of Dynamical Objects indicated by different sets of potential genes which are getting actualized. The semiotic analysis developed here suggests that potential genes can be regarded as a kind of ‘tacit’ representational patterns having strings of DNA as their vehicles. A ‘potential Sign’ is something that may be a Sign of an Object to an Interpretant. A potential Sign, therefore, is a Sign which is not involved in effective semiosis (i.e., effective information), in a given time t. Potential information is defined here as a process of communicating a form embodied in an Object to an Interpretant through the mediation of a Sign that could take place in a given moment (see Ch. 4). A potential gene, as a potential Sign, is just one element in semiosis. This means that potential genes, as patterns in DNA, are not and, also, do not carry information. Rather, they are only the first correlates of a triadicdependent process that we define here as information. Potential genes and, therefore, DNA, carry, harbor, convey only the potentiality of a process we call ‘information’.111 Even if potential genes are treated as patterns in DNA, it will still be the case that to have any effect on the cell as an interpreter they must be subordinated to information as a process, an idea which is consistent, generally speaking, with a process philosophical approach (Rescher 1996). It is within such a process perspective that we treat potential genes as potential Signs, a kind of disposition showing propensities for having certain effects on interpreters, i.e., as potential (semiotic) processes, or tendencies. That is, the framework developed here gives privilege to processes and treats propensities or general tendencies as real. In these terms, it can be made consistent with the general picture about genetic information we find in current molecular 111
In this connection, it is interesting to note that the genetic information system and the human symbolic culture are regarded by Jablonka (2002) as the only information systems capable of transmitting ‘latent’ (‘potential’, ‘non-expressed’, ‘non-actualized’) forms of information. Some kinds of epigenetic cellular information systems also can, in her view, sometimes transmit latent information. Given the framework developed in the present book, this restriction of the capacity of communicating potential information to only a subset of biological information systems demands careful appraisal. Nevertheless, we will limit, for the moment, our conclusions to the genetic information system, leaving this issue to be dealt with elsewhere.
210
Genes, Information, and Semiosis
biology and genetics, provided that this picture is reinterpreted within a general process philosophical stance.112 Effective information, in turn, is not carried by and cannot be identified with entities in DNA, but is, as defined here, the very triadic-dependent, semiotic process by means of which a gene can have any effect on a cell. As such a process, it irreducibly involves Signs, Objects, and Interpretants in a dynamical relationship. Notice, moreover, that it is the cell as an interpreter which coordinates the semiotic processes at stake. Biologically speaking, the genetic material does not do things to the cell, but, rather, it is the cell, as an interpreter, that does something with the genetic material, as we argued throughout this book. The semiotic analysis we developed also allows us to offer an interesting account of the ‘transmission’ of information. It is not effective information that is being communicated when we observe, for instance, ‘vertical transmission’, say, from parent to offspring.113 From the perspective of our results, what is being communicated is only potential information, i.e., the potentiality of the process we call information, which can be said, as explained above, to be carried by stretches of DNA. Signs in DNA will only become elements in effective information when interpreted by the cell. Effective information itself cannot be carried from one system to another, but only potential information can be ‘carried’ by the first correlates of triads, Signs (the vehicles of which, in biological systems, are typically physicochemical entities). In this connection, we think Jablonka points in the right direction when she uses the term ‘transmission’ (taking into account Oyama’s criticism of the typical usage of informational terms in biology) “not for the handing over of pre-existing entities, but to denote any process that results in an organization pattern from one entity being reproduced in another. Thus, when talking about heredity, an entity is related to others by special processes that lead to the reconstruction of 112
In this section, we are presenting a more conservative interpretation of the results of our semiotic analysis of the genetic information system, in which ‘information’ is treated as a process, but ‘genes’ are still treated in a manner which is closer to viewing them as ‘entities’ in DNA, in comparison to the more radical interpretation we will present in the next section. 113 Two modes of ‘information transmission’ are usually recognized: ‘horizontal transmission’, between individuals belonging to the same generation, and ‘vertical transmission’, from one generation to another.
Chapter 9. Genes, information, and semiosis
211
its organization in those other entities” (Jablonka 2002: 588–589). Considering the genetic inheritance system, this can be taken to mean, in our terms, that the pre-existent entities that are transmitted from one generation to another carry only potential information. They are, so as to say, ‘fossils’ in DNA, which will become, only when actualized, Signs in action, developmental resources (among others) in a set of processes that will end up reconstructing the form or organization of an entity in other entities, which will be, thus, of the same class of that prior entity. If we consider the communication of forms by genes from an evolutionary perspective, we will be in a position to claim that in this case forms are communicated from Dynamical Objects (functional proteins) to interpreters through genes as Signs, this being the reason why the Dynamical Object is the primary constraining factor in semiosis (see Figure 7). To clarify the matter, suppose, for the sake of the argument, that a stretch of DNA which codes for a sequence of amino acids indicating a functional protein suffers a mutation in time t1 that turns it into a potentially more effective gene, i.e., after that mutation, if that potential gene is actualized, it will indicate a functional protein which is more effective than the previous one, coded for by the wild allele. Suppose also that this protein plays an important adaptive role in a given lineage, affecting the likelihood of survival and successful reproduction of the organisms carrying it in t1. That potential gene will tend to be preserved by natural selection in the future generations of that lineage, in times t2, t3, …, tn, if the selective regimen remains the same in the significant variables affecting the survival and reproduction of the organisms at stake. The form of the Dynamical Object in t1 increases the chance of the Sign indicating it being present in the next generation of interpreters, in t2, in high frequency. The form of the Dynamical Object is communicated to the interpreters in the future generations through the mediation of the Sign. It is in this sense that we can say that form is communicated, from an evolutionary perspective, from a functional protein, as a Dynamical Object, to a gene, as a Sign. Notice that we are not postulating any inversion of the central dogma (as if sequences of amino acids in proteins might determine sequences of nucleotides in DNA). We are referring, rather, to the effect of functional proteins on the likelihood of certain Signs, certain genes, being present in future generations.
212
Genes, Information, and Semiosis
In somatic time scale, in a given generation, the form — as a type114 — which was evolutionarily communicated from a Dynamical Object at t1 — as a token of that type — to the interpreter by the mediation of the Sign is then communicated from the Sign to the Dynamical Object — as a token of the type which was communicated — in t2, through the mediation of the Interpretant. Thus, the interpreter will be able to produce through habits acquired in evolution and development a new token of the Dynamical Object. Examining the process, therefore, from a proximal, rather than a distant (evolutionary) perspective, we can say that effective information is a process by means of which a form is communicated from COMPS (as a potential Sign) to COMPO through COMPI, indicating a Dynamical Object, the final form of which will depend on constraints established by both COMP S (at the micro-semiotic level) and a series of habits, regularities, at work at the macro-semiotic level. When a Dynamical Object, a functional protein, is finally put into action in a cell, its actual effect on the cell, its Dynamical Interpretant, takes place at t2. The presence of that functional Dynamical Object in a given cell at t2 is mediated by the communication of the form through the gene as a Sign, which constrains the future semiotic processes in daughter cells. This constraint on the future semiotic processes increases the likelihood of repeating the successful, adaptive past interactions with the circumstances. A final argument should be offered to support the claim that information should be identified with a process by which a Sign has an effect on an interpreter, and not with any single element of the triadic-dependent process itself. To build this argument, we will consider in turn each element in a triadic-dependent process that might be regarded as information. Consider, first, that the presence of a given Sign S in DNA cannot be information in itself, since S can be present in each and every cell of an organism, even in those in which it is not expressed and, therefore, has no actual effect on the cell.
114
To understand what we mean by ‘type’ here, consider that if we have a protein-token, say, a particular molecule of fibronectin, it is a token of a type, fibronectin. It is because of the communication of a general form, which defines ‘fibronectin’, from one generation to another by Signs in DNA that a particular, a fibronectin-token, is reconstructed in a given generation.
Chapter 9. Genes, information, and semiosis
213
Secondly, consider that the presence of a Dynamical Object in a given cell do have an effect on it, as its Dynamical Interpretant. Nevertheless, the Dynamical Object has an effect on the cell by means of the communication of its form to an Interpretant by the mediation of the Sign. Information lies in the process of communicating a form, and not in the form itself of the Dynamical Object. By the same token, we should not identify information with the Immediate Object, which simply indicates the Dynamical Object. Thirdly, the Immediate Interpretant is the range of interpretability of a Sign, and, thus, it doesn’t have by itself an effect on the interpreter, but is rather a set of habits which allows something to mean something else and, thus, have an effect on the interpreter. Finally, we should consider the possibility that an environmental cue E to which a given cell responds is information. Surely, there are reasonable grounds for claiming that an environmental cue is ‘informative’, and, no doubt, when we focus on the cell as a whole, a cue E to which the cell responds is involved in a process by which it has an effect on the interpreter (the cell). On then grounds of the framework developed here, we call this process ‘information’. However, by exploring this intuition further, and focusing on the relation between E and changes in gene expression as a subset of possible responses, we can see that when a cell answers to E by changing its pattern of gene expression, potential Signs in its DNA are actualized, allowing the cell to answer to E. That is, the cell can retrieve, given the genetic mechanisms of transmission of potential Signs, past successful, adaptive interactions with environmental cues that have the same character as E in evolutionary events that happened in previous generations. But this means, then, that the cellular system operates to answer to a difference in its environment by changing its internal states. We should still look, in this case, inside the cellular system in order to find the change which takes place when it answers to an environmental cue. We are back, then, at the focal level of our analysis, and, at this level, we already discarded Signs in DNA, Immediate Objects, Immediate Interpretants, Dynamical Objects, and Dynamical Interpretants as possible single ‘bearers’ or ‘units’ of information. We reach, as a conclusion, the same idea we have been advocating throughout this book, namely, that genetic information is a triadicdependent, semiotic process.
214
Genes, Information, and Semiosis
2. What is a gene? A more radical interpretation In the previous section, we basically presented a more conservative way of interpreting the outcomes of the semiotic analysis developed in this book. In this account, ‘information’ is treated as a process, but ‘genes’ are conceived in a manner which is closer to viewing them as ‘entities’ in DNA. This interpretation is in line with a concept Griffiths and Neumann-Held (1999) named ‘the contemporary molecular gene concept’. In the end of Ch. 2, we advanced that our analysis would lead to a pair of alternative interpretations. One of them is this more conservative interpretation, in which genes can be treated as sets of domains in DNA, as Fogle (1990, 2000) proposes, but these sets are not regarded as carriers of genetic information (contra Fogle), since information is defined here as a process. We also consider that sets of domains are built by the cellular system, as a combination of nucleic acid sequences corresponding to an RNA or polypeptide product (Guimarães, Pardini 1992). And we should also take into account that the epistemic practices of communities of researchers reconstruct these sets of domains as epistemic objects (Rheinberger 2000) which operate as targets of research. In a second, more radical interpretation, genes themselves can be treated as the entire molecular process underlying the capacity to express a particular product. This account is basically similar to Neumann-Held’s ‘process molecular gene concept’ (Neumann-Held 1999, 2001, 2006; Griffiths, Neumann-Held 1999). Clearly, process molecular gene concepts are also reconstructed as epistemic objects by communities of researchers. We do not see the possibility of cashing out two distinct interpretations of our analysis as a problem. As we stressed in Ch. 2, we do not think it is either necessary or desirable to have a single definition or understanding of ‘gene’. The more radical interpretation has a number of advantages, since it highlights the basic idea that a stretch of DNA only has any effect on a cell when expressed. Otherwise, it is a rather inert material. A stretch of DNA can be interpreted differently, depending on the cell type, or the particular state of a cell of a given type, or the age of the
Chapter 9. Genes, information, and semiosis
215
cell, or a specific developmental state, or a different species, ‘to the point of promiscuity’ (Guimarães, Moreira 2000).115 Nevertheless, this process approach to genes faces important problems, which should be taken in due account. In Ch. 2, we addressed some troublesome consequences of Neumann-Held’s process molecular gene concept pointed out by Moss (2001). Let us briefly come back to them here. First, in view of phenomena such as alternative splicing, this concept would substantially increase the number of genes in eukaryotes. Second, as the multimolecular systems associated with transcription and splicing would have to be included in it, the process molecular gene would jump to a higher level in the biological hierarchy. Third, it would be hard to individuate process molecular genes, given the extreme context-dependence of gene expression. We will begin with the third problem. Suppose that the scientific community eventually comes to the conclusion that genes are processes. Then, it would be simply the case that a system to classify process molecular genes would have to be devised, so as to circumvent the problem of their context dependence. After all, we cannot deny that what a stretch of DNA means to a cell dramatically varies on the dependence of context, to use once again Guimarães and Moreira’s metaphor, it varies to the point of promiscuity. Therefore, the process molecular gene concept doesn’t create a problem for the taxonomy of genes, as Moss argues, but it simply highlights a problem which is inherent in biological systems, or, to put it differently, it merely points out to a difficulty we cannot and should not ignore. Furthermore, the problem of individuating processes looks just natural when one adopts a process philosophical approach, as Neumann-Held does, and is committed to a criticism of the ‘substance paradigm’ or the ‘myth of the substance’, as Seibt (1996) calls it. In a philosophical account which treats processes as more fundamental than entities, as we find in thinkers such as Alfred N. Whitehead, Charles S. Peirce, Charles Hartshorne, Paul Weiss, Samuel Alexander, Conway Lloyd Morgan, and Andrew Paul Ushenko (see Rescher 1996, 2002), it is an expected outcome that one will face problems such as, for instance, that of developing a taxonomy of highly contingent ontological cate115
Guimarães and Moreira quote in this passage Futuyma’s (1992) remarkable saying that DNA only makes sense in the context of its functional occurrence, i.e., inside the biological system. Otherwise, it is only a molecule, neither more nor less interesting than bones (fossils?) or flowers.
216
Genes, Information, and Semiosis
gories, as processes are. Indeed, processes at first seem harder to individuate than entities. But should we shy of developing such an account because we anticipate that such problems will appear? This doesn’t seem a good decision. We should rather develop the approach and, when problems such as that of building a manageable taxonomy of genes as processes appears, we should attempt to find ways to solve it. Otherwise, we will simply give up before doing the job! We should acknowledge the difficulty, that’s sure, but, once acknowledged, we should develop tools for addressing the problem of building a proper taxonomy for genes conceived as processes. As to the first problem, a similar argument applies. If the result of a proper account of genes is an increase in the number of genes in eukaryotes, does it really matter? Suppose that a process approach turns out to be more adequate in view of our current knowledge about genes — something we concede it’s still to be demonstrated. Then, as in the case of the first problem, it would make no sense to give up this approach simply because the number of genes would increase. Furthermore, the process molecular gene concept can arguably help us address problems in the domain of gene counting, such as Claverie’s (2001) N value paradox. While the C value paradox was grounded on the observation that eukaryote haploid DNA content (C value) was unrelated to organismic complexity or to the number of protein coding genes (see Cavalier-Smith 1978), the N value paradox is related to the issue that the number of genes (’gene content’, as Claverie writes) also does not correlate well with our intuitive perception of organismal complexity. This is related to the fact that the number of human genes estimated by the Human Genome Project was far below previous estimates, which ranged from 50,000 to more than 140,000 genes. While the public effort reported 30,000–40,000 (Lander et al. 2001), the private endeavor identified 26,000–38,000 transcriptional units (Venter et al. 2001). These numbers are rather close, for instance, to the gene content of Drosophila melanogaster (14,000 genes), Caenorhabditis elegans (ca. 19,000), and Arabidopsis thaliana (25,498). The number of human proteins, however, is estimated in 90,000 (Magen, Ast 2005). This is the basis for the paradox: if each gene coded for a single protein (as the unit concept implies, see Ch. 2), how could one explain, say, that the human proteome is about five times bigger than that of the fruit fly? After all, the difference in gene content doesn’t match this difference in proteome.
Chapter 9. Genes, information, and semiosis
217
Obviously, a solution to the N value paradox comes from the mechanisms responsible for proteome expansion in metazoans, such as alternative splicing (Maniatis, Tasic 2002). Nevertheless, we should pay attention to some caveats added by Claverie (2001). He warns that part of the problem results from the lack of an adequate and clear measurement of what we mean by ‘biological complexity’. The value of 30,000 human genes will come as a surprise depending on how one measures complexity. The lack of a clear and unambiguous definition of complexity, let alone a measurement, has been long recognized as a problem (see, e. g., Adami 2002). Claverie has his own solution to the problem. He declares: “I personally favor defining the complexity of an organism as the number of theoretical transcriptome states that its genome could achieve, where the transcriptome represents the universe of transcripts for the genome” (Claverie 2001: 1255). He considers that the estimates resulting from this account “… illustrate how a relatively small number of genes could be sufficient to generate a tremendous biological complexity”, and are, also, […] consistent with the common view that biological sophistication evolves through the development of more individually and finely regulated gene expression mechanisms, rather than a sheer increase in the number of genes. (Claverie 2001: 1255)
These are sound arguments. Notice that we can take a reference to gene expression mechanisms as a way to solve the N value paradox as a possible indication that a process molecular gene approach can be more adequate to understand the relationship between genome and proteome. After all, the mechanisms which lead to proteome expansion in metazoans would be, in this account, simply included in genes. The number of process molecular genes would increase, as a result of the diversity and context-dependence of gene expression mechanisms, and this increase would make them fit better into proteome estimates. We can see that an increase in the number of genes comes here as an advantage rather than a shortcoming, in contrast with Moss’ (2001) opinion. If we address now a question posed by Szathmáry et al. (2001), we will be able to reinforce this argument for process molecular genes. These authors ask: Can genes explain biological complexity? According to them, a measure of complexity better correlated with morphology and behavior can be developed on the grounds of the
218
Genes, Information, and Semiosis
connectivity of regulatory networks, involving both transcription factors and the genes they regulate, rather than by sheer number of genes or interactions between genes. If we take Neumann-Held’s definition of process molecular genes, it is clear that the number of these genes would offer an adequate measurement of complexity, in accordance with Száthmary’s et al. proposal, since it would include the very regulatory networks. But this leads us, in turn, to the second problem raised by Moss: transcription, splicing, and other processes involved in gene expression are accomplished by quite complex multimolecular systems. Therefore, to include these processes in genes, as Neumann-Held’s concept requires, would make genes jump to a higher level in the biological hierarchy. Indeed, we think this is a consequential problem, making process molecular genes look less intuitive than they seem at first. Nevertheless, one way to answer this hesitation would be to note, first, that — as it should be clear from the perspective of a philosophical and historical approach to science — the alternative semiotic model developed in this book involves other ontological assumptions when compared to an entity-based model, in the attempt to grasp a dynamics of processes at different levels of complexity, and, as such, the ontological assumptions of this separate model may imply other senses of what is intuitive. Secondly, one might note that the language game of defining concepts in biology often takes for granted, or has as an implicit background, an organicist (and thus emergentist, cf. Gilbert, Sarkar 2000; El-Hani, Emmeche 2000) perspective, in which a concept at one level, say, a biomolecule, is internally related to concepts at other levels, such as cells and organisms. This will make the fact that to have genes is conceptually dependent upon higher levels in the biological hierarchy appears as a natural consequence of the complex nature of biological systems within an emergentist systems perspective. However, in the context of the semiotic analysis developed here, the problem just raised comes along with a further difficulty: we treated information above as a semiotic process irreducibly involving Signs in DNA, Objects, and Interpretants in a dynamical relationship. If we now also treat genes as processes, wouldn’t it be the case that our analysis would be simply conflating genes and information? An answer to this is to point out, first, the fact that information, once defined as in Ch. 4, is a very general concept, that have many specific
Chapter 9. Genes, information, and semiosis
219
applications inside as well as outside biology, but even applying the general concept in a genetic context as here does not commit one to the view that the two notions are synonymous, or that there cannot be other forms of biological information, which is obviously not true. Secondly, that, if it is true that an interpretation of genes as processes faces the difficulties pointed out by Moss, particularly the latter one, it is also true that the idea that molecular genes or, for that matter, genes-D can be treated as entities in DNA does not face lesser difficulties. For instance, we should question whether the segmentation problem can be really solved. It may be the case that all the complexities involved in the expression and regulation of DNA segments make any attempt to divide chromosomes into genes in an appropriate way doomed to failure. In this case, it will be impossible, in the end, to identify genes with particular segments on chromosomes, while an understanding of genes as processes can avoid this individuation problems one faces when treating genes as entities. It would be a rash decision to cast away Neumann-Held’s molecular process gene concept on the grounds of the problems discussed above. After all, the treatment of genes as entities in DNA seem to face problems which are as serious as (if not more critical than) those faced by a process-oriented view of genes. In fact, the resistance to the claim that genes can be treated as processes which go beyond DNA can be traced back to the idea that genes, no matter how we understand them, are segments of DNA, which is deeply entrenched in our view about that theoretical concept. Nevertheless, the difficulties faced when we try to segment chromosomes into genes may be taken as suggesting that this substance-view of genes, even though deeply entrenched, can be wrong. In a suggestive way, Keller writes: […] virtually every biologically significant property conventionally attributed to the DNA — including its stability — is in fact a relational property, a consequence of the dynamic interactions between DNA and the many protein processors that converge upon it. The very meaning of any DNA sequence is relational — for the purpose of understanding development or disease, the patterns of genetic expression are what really matters, and these patterns are under the control of a vastly complex regulatory apparatus, and they cannot be predicted from knowledge of the sequence alone. (Keller 2005: 4)
220
Genes, Information, and Semiosis
Neumann-Held’s process molecular gene concept is favored by the recognition of this dependence of DNA on expression processes and cell context to make any difference to living systems. Keller also argues that “genes, by definition, do not have meaning in isolation”, and treats the cell as “a meaning making system that turns nucleotide sequences into genes” (Keller 2005: 9). In these terms, DNA sequences are turned into genes by the cell’s interpretative processes, as our semiotic analysis of the genetic information system stresses. Keller also seems to derive from her arguments a tendency towards a process approach to genes: she writes that we have been trying to build “a biology out of nouns, a science constructed around entities”, and suggests that “it is time for a biology built out of verbs, a science constructed around processes”, or, else, “a conceptual framework that rests on a dynamical and relational epistemology” (Keller 2005: 9). In her view, genes can be revived for the 21st century in a new science of biology, but only if they are reconceptualized as verbs. This is indeed what the radical interpretation of our semiotic analysis of genes entails: genes would be processes, rather than entities or even sequences of entities in DNA. This interpretation leads us away from the ideas that genes might be structural and/or functional units in DNA, into new ways of thinking about biological function, in which function is not to be found in particular genes, nor in the structure of DNA and its protein products, but rather “in the communication networks of which the DNA and the proteins are part” (Keller 2005: 9). Biology can be conceptualized, thus, as a communication and information science, as it is claimed in the new wave of ‘systems biology’ (see, e. g., Ideker et al. 2001. But cf. comments in Ch. 2 of this book). We can now ask (in an explicitly provocative manner): What is this new conceptualization of biology, if not biosemiotics? In sum, it is true that Moss (2001) pointed out important problems to the process approach to genes. But it is not the case, we think, of simply giving up in the face of these problems. While we shouldn’t at all neglect the problems Moss raised, we shouldn’t also be discouraged by them. As a final word on this issue, we should emphasize we are in favor of a process approach to the genetic information system, much in the same spirit as Keller. The only doubt is whether this process approach should be restricted to genetic information, or be extended to genes themselves. We hope that the above discussion made it clear some arguments for and against this latter move.
Chapter 9. Genes, information, and semiosis
221
3. Concluding remarks We developed in this book an analysis of the genetic information system which is, in our view, in full accordance with Peirce’s theory of Signs and his general process approach to philosophy. Consider, for instance, the claim that potential genes carry only the potentiality of information interpreted as a process rather than an entity, and, accordingly, that information, not even in a potential sense, corresponds to sequences of nucleotides in stretches of DNA. This idea can be straightforwardly related, in turn, to the claim that a sequence of nucleotides has no intrinsic meaning in the absence of a cell to interpret it, as argued above. Furthermore, DNA becomes effective information only when it is used as ‘data’ (Atlan, Koppel 1990) (or, as we prefer, Signs) by an active and complex system of interpretation in the cell, i.e., when potential genes are actualized in response to intraand/or extracellular signals. A nucleotide sequence means nothing apart from the dynamics of the cell. These ideas are exquisitely consistent with Peirce’s claim that ‘... it is impossible to deal with a triad without being forced to recognize a triad of which one member is positive but ineffective, another is the opponent of that, a third, intermediate between these two, is allpotent’ (CP 4.317). Signs in DNA, potential genes, can be understood, if we borrow Peirce’s terms, as being ‘positive but ineffective’. Indeed, DNA, in a cell system, cannot do anything by itself, while the cell, in turn, can do things with DNA, by actualizing potential Signs in it, so as to indicate some useful molecule, say, a functional protein (an ‘opponent’), a Dynamical Object, which determines a third, a Dynamical Interpretant, which is ‘all-potent’, playing a given function or resulting in a given dysfunction, depending on the specific case at stake. Furthermore, the account developed here is also in accordance with the general picture of genes and how they are expressed in genetics and molecular biology, with the fundamental difference that information is differently conceptualized, since it is treated as a process, and not as something DNA can contain, harbor, carry, etc. This is particularly the case of the more conservative interpretation of our results. In the more radical interpretation, not only information but also genes are treated as processes, and, consequently, the tension between this interpretation and the standard picture of genes as
222
Genes, Information, and Semiosis
stretches of DNA makes the semiotic account of the genetic information system developed in this book less compatible with the general framework of genetics and molecular biology. Nevertheless, the more radical interpretation of the outcomes of our analysis is still reasonably consistent with this latter framework. We consider the compatibility of our analysis with the framework of genetics and molecular biology as a strength rather than a weakness. After all, it can be taken to mean that it is possible to develop in this way a more consistent understanding of ‘information’ which can be smoothly integrated into established knowledge in that framework. Nevertheless, it is open to investigation an evaluation of the pros and cons of building a semantic/pragmatic concept of information inside current paradigms in genetics and molecular biology, or of striving for promoting a conceptual revolution in these disciplines from the standpoint of biosemiotics. One or the other project will seem attractive for different groups of researchers. We hope we have been successful in showing how the conceptual and methodological tools offered by biosemiotics can help us make it more precise what is information in biological systems. If we succeeded in this point, this work can be seen as a contribution to the argument for an explanatory role for biosemiotics with regard to fundamental notions in biology which are related to a communicational and informational vocabulary, as the concept of information. It is never too much to remind that we don’t have an established notion of biological information up to this point, and it is a basic contention of this work that biosemiotics can help in building such a notion. In this way, we are more likely to go beyond the unfortunate situation of information talk in biology as a loose bunch of metaphors with no clear meaning. We hope our arguments have shown how biosemiotics can contribute to the project of building a theory about information in biology, including both semantic and pragmatic dimensions of information. Rather than eliminating information talk from biology, the biosemiotic point of view regards it as essential to a way of conceiving biological systems that grasp their fundamental difference from standard chemical and physical systems. It is just the case of employing adequate tools, such as those offered by Peirce’s theory of
Chapter 9. Genes, information, and semiosis
223
Signs, to clarify the nature of information in a living system.116 Consider, for instance, Oyama’s (2000) argument that, while genes are represented as if they contained information about how an organism will develop, they will continue to be treated as determining causes, lending support to genetic determinism. The notion of information arising from the biosemiotic analysis presented here suggests that the problem is not really in the notion of information in itself, but rather in the way information has been typically conceived in genetics and molecular biology, namely, in such a way that it was reduced to merely sequential information in a string of DNA. If we consider, as above, that strings of DNA only contain potential genes; that potential information, although arguably carried by stretches of DNA, should be treated as the potentiality of a process; and effective information is a triadic-dependent process including not only Signs in DNA but also Objects and Interpretants, we will be in a much better position to picture information as being fundamentally dynamical and distributed, being related not only to genes but to any structure that can act as a Sign. Maybe, we can even follow Keller (2000: 146) in her suggestion of a cellular program that is not limited to DNA, but is rather a shared program in which all cell components functions alternatively as ‘instructions’ and ‘data’ (or, as we prefer, Signs). Williams (1997: 476–477), in his summary of a symposium promoted by Ciba Foundation to discuss the future of the reductionist approach in biology, states that the biggest challenge to reductionism comes from the concept of information.117 Some participants of that meeting — he reports — “... felt that a deeper understanding of the role of information may yet throw a spanner in the grand reductionist scheme.” In a sense, he argues, information in biological systems is “fully consistent with” reductionist principles of physics and chemistry, because it is “carried and received by molecules.” In terms of a biosemiotic analysis, these molecules that ‘carry’ and ‘receive’ ‘information’ are regarded as just potential Signs. These potential Signs, however, to be effectively ‘informational’ should be part of 116
But, surely, we don’t intend to argue that only Peirce’s theory of Signs offers invaluable tools for a semiotic analysis of biological systems and processes. There are other tendencies in biosemiotics and they can certainly bring interesting contributions to our understanding of life. 117 More recently, it has been even claimed that biology is the science of information (Ideker et al. 2001).
224
Genes, Information, and Semiosis
triadic-dependent, semiotic processes, involving more than just these potential entities. The conclusion that information is a process rather than an entity shows how a careful analysis of what is information in biological systems, based on a coherent framework such as that of semiotics, can indeed overcome a one-sided reliance on reductionist approaches to biology. According to the picture presented in this book, the meaning of a gene is highly context-sensitive. After all, information is highly context-sensitive, and genes can only mean something by being Signs within a triadic-dependent process defined here as information. A Peircean approach to the concepts of gene and information entails that both should be seen in the contexts in which information is handled by an interpreter, a conclusion which is in accordance with ideas stressed by a number of authors involved in the debates about these concepts (e. g., Nijhout 1990, Keller 2000, Hall 2001, Jablonka 2002), and highlights the pragmatic dimension of genetic information. The meaning of a gene is not contained in the sequence of nucleotides in a string of DNA, but rather emerges as a process involving the larger system by which genes are interpreted. Notice, furthermore, that both genetic information and genes (no matter if they are seen as processes or patterns in DNA) are seen in our account as constructed, instead of only given. They are partly built as epistemic objects through a set of practices of investigation and predication elaborated and employed by the community of researchers who work with genes and information (see Rheinberger 2000). But the system in which genes and genetic information acquire a meaning also plays an important role in building them. As Pardini and Guimarães (1992) propose, in his systemic concept of the gene, the gene, for instance, can be seen as a combination of (one or more) nucleic acid (DNA or RNA) sequences which is established by the system which interprets them, and, thus, give them a specific meaning. This system includes the whole cell and also its peri- and extracellular matrices (Hall 2001). As Hall (2001) emphasizes, one of the major unresolved problems in biology is how to place genes in context. His answer is in agreement with our conclusions in this paper: ‘Simply, the gene’s home, context, and locus of operation is the cell’.118 This may seem at first 118
In fact, as Hall (2001: 228) also recognizes, genes have multilevel homes or contexts, given the nested structure of living systems. Cells are given preeminence for their widely acknowledged role as fundamental units of organic structure and
Chapter 9. Genes, information, and semiosis
225
sight too obvious to be of any relevance. Nevertheless, as Hall stresses, we have been slow to recognize that the cell is not only the place where genes reside, but “the cell enables the gene, allowing it to play its role(s) in development and evolution” (Hall 2001: 226). The very fact that the mainstream representation of genes is such that all information in a cell end up being deposited in DNA shows how slow we have been to recognize that DNA is enabled by the cell to perform the roles it performs. As a consequence of the semiotic analysis offered in this book, the interpretation of what is information in a cell system and, in particular, of how potential Signs in DNA can be actualized so as to be part of effective information, clearly ascribes to the cell, as an interpreter of Signs in DNA, the capacity of enabling genes, much in the sense proposed by Hall (2001). Cells enable DNA to perform its roles by harnessing the behavior of this macromolecule so as to make it operate in a particular way that is demanded by a given environmental situation a cell faces at certain locations and times. The mechanisms that allow cells to constrain the operation of DNA to their own needs involve the establishment of boundary conditions by a macro-semiotic level of signalling pathways, as discussed above. They show that DNA molecules are governed by the cell, rather than command the cell in a dictatorial way, as the metaphors of genetic ‘programs’ and ‘controllers’ suggest. Biological systems function by means of a ‘democratic’ rather than a ‘dictatorial’ control structure, i.e., there is neither genomic nor metabolic supremacy over other cellular processes (Bruggeman et al. 2002). The recognition that the cell is the context of genes unravels new and difficult challenges, for instance, that of identifying and understanding the spatial and temporal contexts (often quite complex and multifaceted) in which genes operate, or, as Hall (2001: 228) puts it, ‘to unravel the epigenetic code underlying developmental evolution’, which is far more complicated than the genetic code of four ‘letters’ arranged in groups of three. After all, it encompasses a whole array of genetic and non-genetic factors that ultimately enables genotype to function as a semiotic ressource in the complicated developmental emergence of all the diverse characteristics of an organism’s phenotype. We should leave, however, the analysis of other biological features relevant to the project of a semiotic analysis of the function, and, also, for the fact that there are an enormous number of (unicellular) organisms which have no level higher than the cell.
226
Genes, Information, and Semiosis
genetic information system for subsequent works. The arguments developed in this book are, in our view, sufficient to show both the relevance and the far-reaching consequences of such an analysis.
References Adami, Christoph 2004. Information theory in molecular biology. Physics of Life Reviews 1: 3–22. Adams, Frederick 2003. The informational turn in philosophy. Minds and Machines 13: 471–501. Alberts, Bruce; Johnson, Alexander; Lewis, Julian; Raff, Martin; Roberts, Keith; Walter, Peter 2002. Molecular Biology of the Cell. (4th ed.) New York: Garland Science. Andersen, Peter Bøgh; Finnemann, Niels Ole; Christiansen, Peder Voetmann; Emmeche, Claus (eds.) 2000. Downward Causation: Minds, Bodies and Matter. Aarhus: Aarhus University Press. Ashburner, Michael; Ball, Catherine A.; Blake, Judith A.; Botstein, David; Butler, Heather; Cherry, J. Michael; Davis, Allan P.; Dolinski, Kara; Dwight, Selina S.; Eppig, Janan T.; Harris, Midori A.; Hill, David P.; Issel-Tarver, Laurie; Kasarskis, Andrew; Lewis, Suzanna; Matese, John C.; Richardson, Joel E. (The Gene Ontology Consortium) 2000. Gene ontology: Tool for the unification of biology. Nature Genetics 25: 25–29. Atlan, Henry; Koppel, Moshe 1990. The cellular computer DNA: program or data? Bulletin of Mathematical Biology 52(3): 335–348. Azzone, G. F. 1998. The cement of medical thought: Evolutionary emergence and downward causation. History and Philosophy of the Life Sciences 20(2): 163– 187. Baas, Nils A. 1996. A framework for higher-order cognition and consciousness. In: Hameroff, S. R.; Kaszniak, A. W.; Scott, A. C. (eds.), Toward a Science of Consciousness. Cambridge: MIT Press, 633–648. Baas, Nils A.; Emmeche, Claus 1997. On emergence and explanation. Intellectica 2(25): 67–83. [= SFI Working Paper 97-02-008. Santa Fe Institute, New Mexico] Bailey, Andrew 1999. Supervenience and physicalism. Synthese 117: 53–73. Barabási, Albert-László; Oltvai, Zoltán N. 2004. Network biology: Understanding the cell’s functional organization. Nature Reviews, Genetics 5: 101–113. Barbieri, Marcello 2003. The Organic Codes: An Introduction to Semantic Biology. Cambridge: Cambridge University Press. Bateson, Gregory 1972. Steps to an Ecology of Mind. New York: Ballantine Books. Beadle, George W.; Tatum, Edward L. 1941. Genetic control of biochemical reactions in Neurospora. Proceedings of the National Academy of Sciences of the United States 27: 499–506. Beckermann, A.; Flohr, H.; Kim, J. (eds.) 1992. Emergence or Reduction? Essays on the Prospects of Nonreductive Physicalism. Berlin: Walter de Gruyter. Bedau, Mark A. 1997. Weak emergence. Noûs 31(Suppl. Philosophical Perspectives 11): 375–399. — 2002. Downward causation and autonomy of weak emergence. Principia 6(1): 5–50.
228
Genes, Information, and Semiosis
Benzer, Seymour 1957. The elementary units of heredity. In: McElroy, W.; Glass, B. (eds.), The Chemical Basis of Heredity. Baltimore: John Hopkins Press, 70–93. Bergandi, Donato 1995. “Reductionist holism”: an oxymoron or a philosophical chimera of E. P. Odum’s systems ecology? Ludus Vitalis: Journal of Philosophy of Life Sciences 3: 145–180. Bergman, Mats 2000. Reflections on the role of the communicative sign in semeiotic. Transactions of the Charles S. Peirce Society 36(2): 225–254. Bertalanffy, Ludwig von 1973. General System Theory: Foundations, Development, and Applications. London: Penguin. Bickhard, Marc H.; Campbell, Donald T. 2000. Emergence. In: Andersen et al. 2000: 13–34. Black, Douglas L. 2003. Mechanisms of alternative pre-messenger RNA splicing. Annual Review of Biochemistry 72: 291–336. Blitz, David 1992. Emergent Evolution: Qualitative Novelty and the Levels of Reality. Dordrecht: Kluwer. Boogerd, F. C.; Bruggeman, F. J.; Richardson, R. C.; Stephan, A.; Westerhoff, H. V. 2005. Emergence and its place in nature: A case study of biochemical networks. Synthese 145(1): 131–164. Brenner, Sidney 1999. Theoretical biology in the third millennium. Philosophical Transactions of the Royal Society of London B 354: 1963–1965. Brier, Søren 1996. From second-order cybernetics to cybersemiotics: A semiotic re-entry into the second-order cybernetics of Heinz von Foerster. Systems Research 13: 229–244. Britten, Roy J.; Kohne, David E. 1968. Repeated sequences in DNA. Science 161: 529–540. Broad, C. D. 1919. Mechanical explanation and its alternatives. Proceedings of the Aristotelian Society 19: 86–124. — 1925. The Mind and its Place in Nature. London: Kegan Paul, Trench, Trubner & Co. Brosius, Jürgen; Gould, Stephen Jay 1992. On ‘genomenclature’: a comprehensive (and respectful) taxonomy for pseudogenes and other ‘junk DNA’. Proceedings of the National Academy of Sciences of the United States 89(22): 10706–10710. Brown, Terence A. 2002. Genomes. (2nd ed.) Oxford: BIOS Scientific Publishers. Bruni, Luis Emilio 2003. A Sign-theoretic Approach to Biotechnology. Ph.D. dissertation. Institute of Molecular Biology, University of Copenhagen. — 2007. Cellular semiotics and signal transduction. In: Barbieri, Marcello (ed.), Introduction to Biosemiotics: The New Biological Synthesis. Dordrecht: Springer, 365–407. — 2008. Gregory Bateson’s relevance to current molecular biology. In: Hoffmeyer, Jesper (ed.), A Legacy for Living Systems: Gregory Bateson as Precursor to Biosemiotics. Dordrecht: Springer, 93–119.
References
229
Brunning, Jackeline 1997. Genuine triads and teridentity. In: Houser, Nathan; Roberts, Don D.; Evra, James van (eds.), Studies in the Logic of Charles Sanders Peirce. Bloomington: Indiana University Press, 252–270. Brunning, Jackeline; Forster, Paul (eds.) 1997. The Rule of Reason: The Philosophy of Charles Sanders Peirce. Toronto: University of Toronto Press. Bunge, Mario 1977. Levels and reduction. American Journal of Physiology 233(3): R75–R82. Burch, Robert W. 1991. A Peircean Reduction Thesis. Lubbock: Texas Tech University Press. — 1997. Peirce’s reduction thesis. In: Houser, Nathan; Roberts, Don D.; Evra, James van (eds.), Studies in the Logic of Charles Sanders Peirce. Bloomington: Indiana University Press, 234–251. Burian, Richard M. 1985. On conceptual change in biology: The case of the gene. In: Depew, David J.; Weber, Bruce H. (eds.), Evolution at a Crossroads: The New Biology and the New Philosophy of Science. Cambridge: MIT Press, 21– 24. Camargo, Anamaria A.; Samara, Helena P. B.; Dias-Neto, Emmanuel; Simão, Daniel F.; Migotto, Ítalo A.; Briones, Marcelo R. S.; Costa, Fernando F.; Nagai, Maria Aparecida; Verjovski-Almeida, Sergio; Zago, Marco A.; Andrade, Luis Eduardo C.; Carrer, Helaine; El-Dorry, Hamza F. A.; Espreafico, Enilza M.; Habr-Gama, Angelita; Giannella-Neto, Daniel; Goldman, Gustavo H.; Gruber, Arthur; Hackel, Christine; Kimura, Edna T.; Maciel, Rui M. B.; Marie, Suely K. N.; Martins, Elizabeth A. L.; Nóbrega, Marina P.; Paçó-Larson, Maria Luisa; Pardini, Maria Inês M. C.; Pereira, Gonçalo G.; Pesquero, João Bosco; Rodrigues, Vanderlei; Rogatto, Silvia R.; Silva, Ismael D. C. G. da; Sogayar, Mari C.; Sonati, Maria de Fátima; Tajara, Eloiza H.; Valentini, Sandro R.; Alberto, Fernando L.; Amaral, Maria Elisabete J.; Aneas, Ivy; Arnaldi, Liliane A. T.; Assis, Angela M. de; Bengtson, Mário Henrique; Bergamo, Nadia Aparecida; Bombonato, Vanessa; Camargo, Maria E. R. de; Canevari, Renata A.; Carraro, Dirce M.; Cerutti, Janete M.; Côrrea, Maria Lucia C.; Côrrea, Rosana F. R.; Costa, Maria Cristina R.; Curcio, Cyntia; Hokama, Paula O. M.; Ferreira, Ari J. S.; Furuzawa, Gilberto K.; Gushiken, Tsieko; Ho, Paulo L.; Kimura, Elza; Krieger, José E.; Leite, Luciana C. C.; Majunder, Paromita; Marins, Mozart; Marques, Everaldo R.; Melo, Analy S. A.; Melo, Monica Barbosa de; Mestriner, Carlos Alberto; Miracca, Elisabete C.; Miranda, Daniela C.; Nascimento, Ana Lucia T. O.; Nóbrega, Francisco G.; Ojopi, Elida P. B.; Pandolfi, José Rodrigo C.; Pessoa, Luciana G.; Prevedel, Aline C.; Rahal, Paula; Rainho, Claudia A.; Reis, Eduardo M. R.; Ribeiro, Marcelo L.; Rós, Nancy da; Sá, Renata G. de; Sales, Magaly M.; Sant’anna, Simone Cristina; Santos, Mariana L. dos; Silva, Aline M. da; Silva, Neusa P. da; Silva Jr., Wilson A.; Silveira, Rosana A. da; Sousa, Josane F.; Stecconi, Daniella; Tsukumo, Fernando; Valente, Valéria; Soares, Fernando; Moreira, Eloisa S.; Nunes, Diana N.; Correa, Ricardo G.; Zalcberg, Heloisa; Carvalho, Alex F.; Reis, Luis F. L.; Brentani, Ricardo R.; Simpson, Andrew J. G.; Souza, Sandro J. de 2001. The contribution of 700,000 ORF sequence tags to the definition
230
Genes, Information, and Semiosis
of the human transcriptome. Proceedings of the National Academy of Sciences of the United States 98: 12103–12108. Campbell, Donald T. 1974. ‘Downward causation’ in hierarchically organised biological systems. In: Ayala, Francisco; Dobzhansky, Theodosius (eds.), Studies in the Philosophy of Biology: Reduction and Related Problems. Berkeley: University of California Press, 179–186. Cangelosi, A.; Turner, H. 2002. L'emergere del linguaggio. In: Borghi, A. M.; Iachini, T. (eds.), Scienze della Mente. Bologna: Il Mulino, 227–244. Cariani, Peter 1989. On the Design of Devices with Emergent Semantic Functions. Ph.D. dissertation, Department of Systems Science, State University of New York at Binghamton. Ann Arbor: University Microfilms.119 Carlson, Axel Elof 1966. The Gene: A Critical History. Philadelphia: W. B. Saunders. Caro, Mario De; MacArthur, David (eds.) 2004. Naturalism in Question. Cambridge: Harvard University Press. Carroll, Sean B.; Grenier, Jennifer K.; Weatherbee, Scout D. 2005. From DNA to Diversity: Molecular Genetics and the Evolution of Animal Design. Oxford: Blackwell. Chong, Lisa; Ray, L. Bryan 2002. Whole-istic biology. Science 295: 1661. Clark, M. M.; Karpiuk, P.; Galef Jr., B. G. 1993. Hormonally mediated inheritance of acquired characteristics in Mongolian gerbils. Nature 364: 712. Claverie, Jean-Michel 2001. What if there are only 30,000 human genes? Science 291: 1255–1257. Colapietro, Vincent 1989. Peirce’s Approach to the Self: A Semiotic Perspective on Human Subjectivity. Albany: State University of New York Press. Collins, Francis S.; Green, Eric D.; Guttmacher, Alan E.; Guyer, Mark S. 2003. A vision for the future of genomics research: a blueprint for the genomic era. Nature 422: 835–847. Cover, Thomas M.; Thomas, Joy A. 1999. Information theory. In: Wilson, Robert A.; Keil, Frank C. (eds.), The MIT Encyclopedia of the Cognitive Sciences. Cambridge: MIT Press, 404–406. Coyne, Jerry A. 2000. The gene is dead: Long live the gene. Nature 408: 26–27. CP = Peirce 1931–1935. Crick, Francis H. 1958. On protein synthesis. Symposium of the Society of Experimental Biology 12: 138–163. Csete, Marie E.; Doyle, John C. 2002. Reverse engineering of biological complexity. Science 295: 1664–1669. Cummins, Robert 1975. Functional analysis. Journal of Philosophy 72: 741–764. [Also printed in: Allen, Colin; Bekoff, Marc; Lauder, Georg (eds.) 1998. Nature’s Purposes: Analyses of Function and Design in Biology. Cambridge: MIT Press.] — 2002. Neo-teleology. In: Ariew, Andres; Cummins, Robert; Perlman, Mark (eds.), Functions: New essays in philosophy of psychology and biology. Oxford: Oxford University Press, 157–172. 119
Available from www.cariani.com.
References
231
Cunningham, Bryon 2001. The Reemergence of ‘Emergence’. Philosophy of Science 68 (PSA 2000 Proceedings): S62–S75. Deacon, Terrence 1997. The Symbolic Species: The Co-evolution of Language and the Brain. New York: W.W. Norton & Company. Danen, Erik H. J.; Yamada, Kenneth M. 2001. Fibronectins, integrins, and growth control. Journal of Cellular Physiology 189: 1–13. Dawkins, Richard 1982. The Extended Phenotype. Oxford: W. H. Freeman. — 1989 [1976]. The Selfish Gene. (New ed.). Oxford: Oxford University Press. Debrock, Guy 1996. Information and the metaphysical status of the sign. In: Colapietro, V.; Olshewsky, T. (eds.), Peirce’s Doctrine of Signs — theory, applications, and connections. Berlin: Mouton de Gruyter, 80–89. El-Hani, Charbel Niño 1997. Explicações causais do desenvolvimento: são os genes suficientes? Cadernos de História e Filosofia da Ciência, série 3, 7(1): 121–167. — 2002a. On the reality of emergents. Principia 6(1): 51–87. — 2002b. Uma ciência da organização viva: Organicismo, emergentismo e ensino de biologia. In: Silva Filho, W. J. (ed.), Epistemologia e Ensino de Ciências. Salvador: Arcadia/UCSal, 199–244. — 2004. Os debates sobre a emergência de propriedades: Onde estamos atualmente? In: Rabossi, E. (ed.), La Mente y Sus Problemas: Temas Actuales de Filosofía de la Psicología. Buenos Aires: Catálogos, 249–300. El-Hani, Charbel Niño; Pereira, Antonio Marcos 1999. Understanding biological causation. In: Hardcastle, V. G. (ed.), Where Biology Meets Psychology: Philosophical Essays. Cambridge-MA: MIT Press, 333–356. — 2000. Higher-Level descriptions: Why should we preserve them? In: Andersen et al. 2000: 118–142. El-Hani, Charbel Niño; Emmeche, Claus 2000. On some theoretical grounds for an organism-centered biology: Property emergence, supervenience, and downward causation. Theory in Biosciences 119(3/4): 234–275. El-Hani, Charbel Niño; Queiroz, João 2005a. Modos de irredutibilidade das propriedades emergentes. Scientiae Studia 3(1): 9–41. — 2005b. Downward determination. Abstracta 1(2): 162–192. El-Hani, Charbel Niño; Videira, A. A. P. 2001. Causação descendente, emergência de propriedades e modos causais Aristotélicos. Theoria 16: 301–329. El-Hani, Charbel Niño; Pihlström, Sami 2002a. Emergence theories and pragmatic realism. Essays in Philosophy 3(2).120 — 2002b. A pragmatic realist view of emergence. Manuscrito 25: 105–154. — 2005. Realismo, pragmatismo e emergencia. In: Smith, P.; Silva Filho, W. J. (eds.), Significado, Verdade e Interpretação: Davidson e a Filosofia. São Paulo: Loyola, 199–238. El-Hani, Charbel Niño; Queiroz, João; Emmeche, Claus 2006. A semiotic analysis of the genetic information system. Semiotica 160(1/4): 1–68.
120
Available at http://www.humboldt.edu/~essays/pihlstrom.html.
232
Genes, Information, and Semiosis
El-Hani, Charbel Niño; Arnellos, Argyris; Queiroz, João 2007. Modeling a semiotic process in the immune system: Signal transduction in B-cells activation. tripleC 5(2): 24–36. Emmeche, Claus 1990. Det biologiske informationsbegreb [The concept of information in biology]. Aarhus: Forlaget Kimære. [Phd. dissertation.] — 1991. A semiotical reflection on biology, living signs and Artificial Life. Biology and Philosophy 6: 325–340. — 1994. The Garden in the Machine: The Emerging Science of Artificial Life. Princeton: Princeton University Press. — 1997. Autopoietic systems, replicators, and the search for a meaningful biologic definition of life. Ultimate Reality and Meaning 20(4): 244–264. — 1998. Defining life as a semiotic phenomenon. Cybernetics & Human Knowing 5(1): 3–17. — 2000. Closure, function, emergence, semiosis and life: The same idea? Reflections on the concrete and the abstract in theoretical biology. In: Chandler, Jerry L. R.; Vijver, Gertrudis Van de (eds.), Closure: Emergent Organizations and Their Dynamics. (Annals of the New York Academy of Sciences, volume 901). New York: The New York Academy of Sciences, 187–197. — 2003. Causal processes, semiosis, and consciousness. In: Seibt, Johanna (ed.), Process Theories: Crossdisciplinary Studies in Dynamic Categories. Dordrecht: Kluwer, 313–336. Emmeche, Claus; Hoffmeyer, Jesper 1991. From language to nature — the semiotic metaphor in biology. Semiotica 84(1/2): 1–42. Emmeche, Claus; Køppe, Simo; Stjernfelt, Frederik 1997. Explaining emergence: Towards an ontology of levels. Journal for General Philosophy of Science 28: 83–119. Emmeche, Claus; Køppe, Simo; Stjernfelt, Frederik 2000. Levels, emergence, and three versions of downward causation. In: Andersen et al. 2000: 13–34. Emmeche, Claus; Kull, Kalevi; Stjernfelt, Frederik 2002. Reading Hoffmeyer, Rethinking Biology. (Tartu Semiotics Library 3.) Tartu: Tartu University Press. Enard, Wolfgang; Khaitovich, Philipp; Klose, Joachim; Zöllner, Sebastian; Heissig, Florian; Giavalisco, Patrick; Nieselt-Struwe, Kay; Muchmore, Elaine; Varki, Ajit; Ravid, Rivka; Doxiadis, Gaby M.; Bontrop, Ronald E.; Pääbo, Svante 2002. Intra- and interspecific variationin primate gene expression patterns. Science 296: 340–343. EP2 = Peirce 1998. Epp, Christopher D. 1997. Definition of a gene. Nature 389: 537. Falk, Raphael 1986. What is a gene? Studies in the History and Philosophy of Science 17: 133–173. — 2000. The gene — A concept in tension. In: Beurton, Peter J.; Falk, Raphael; Rheinberger, Hans-Jörg (eds.), The Concept of the Gene in Development and Evolution. Cambridge: Cambridge University Press, 317–348. — 2001. Can the norm of reaction save the gene concept? In: Singh, Rama S.; Krimbas, Costas B.; Paul, Diane B.; Beatty, John (eds.), Thinking about
References
233
Evolution: Historical, Philosophical and Political Perspectives. New York: Cambridge University Press, 119–140. Favareau, Donald 2007. The evolutionary history of biosemiotics. In: Barbieri, Marcello (ed.), Introduction to Biosemiotics: The New Biological Synthesis. Dordrecht: Springer, 1–67. Fetzer, James H. 1988. Signs and minds: An introduction to the theory of semiotic systems. In: Fetzer, James H. (ed.), Aspects of Artificial Intelligence. Dordrecht: Kluwer, 133–161. — 1997. Thinking and computing: Computers as special kinds of signs. Minds and Machines 7: 345–364. Fire, Andrew 1999. RNA-triggered gene silencing. Trends in Genetics 15(9): 358–363. Fitzgerald, John 1966. Peirce’s Theory of Signs as Foundation for Pragmatism. The Hague: Mouton & Co. Fogle, Thomas 1990. Are genes units of inheritance? Biology and Philosophy 5: 349–371. — 2000. The dissolution of protein coding genes. In: Beurton, Peter J.; Falk, Raphael; Rheinberger, Hans-Jörg (eds.), The Concept of the Gene in Development and Evolution. Cambridge: Cambridge University Press, 3–25. Freadman, Anne 2004. The Machinery of Talk: Charles Peirce and the Sign Hypothesis. Stanford: Stanford University Press. Freeman, Eugene 1983. The Relevance of Charles Peirce. La Salle: Monist Library of Philosophy. Gabius, Hans-Joachim 2000. Biological information transfer beyond the genetic code: The sugar code. Naturwissenschaften 87(3): 108–121. Gavin, Anne-Claude; Bösche, Markus; Krause, Roland; Grandi, Paola; Marzioch, Martina; Bauer, Andreas; Schultz, Jörg; Rick, Jens M.; Michon, Anne-Marie; Cruciat, Cristina-Maria; Remor, Marita; Höfert, Christian; Schelder, Malgorzata; Brajenovic, Miro; Ruffner, Heinz; Merino, Alejandro; Klein, Karin; Hudak, Manuela; Dickson, David; Rudi, Tatjana; Gnau, Volker; Bauch, Angela; Bastuck, Sonja; Huhse, Bettina; Leutwein, Christina; Heurtier, Marie-Anne; Copley, Richard R.; Edelmann, Angela; Querfurth, Erich; Rybin, Vladimir; Drewes, Gerard; Raida, Manfred; Bouwmeester, Tewis; Bork, Peer; Seraphin, Bertrand; Kuster, Bernhard; Neubauer, Gitte and Superti-Furga, Giulio 2002. Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415: 141–147. Gelbart, William 1998. Databases in genomic research. Science 282: 659–661. Gelder, Tim Van 1998. The dynamical hypothesis in cognitive science. Behavioral and Brain Sciences 21: 615–665. Gilbert, Scott F.; Sarkar, Sahotra 2000. Embracing complexity: Organicism for the 21st century. Developmental Dynamics 219: 1–9. Gillett, Carl 2002. Strong emergence as a defense of non-reductive physicalism: A physicalist metaphysics for ‘downward’ determination’. Principia 6(1): 89–120. Godfrey-Smith, Peter 1999. Genes and codes: Lessons from the philosophy of mind? In: Hardcastle, Valerie Gray (ed.), Where Biology Meets Psychology: Philosophical Essays. Cambridge: MIT Press, 305–331.
234
Genes, Information, and Semiosis
— 2000. Information, arbitrariness, and selection: Comments on Maynard Smith. Philosophy of Science 67(2): 202–207. Gomes, Antonio; Gudwin, Ricardo; Queiroz, João 2003. On a Computational Model of Peircean Semiosis. In: Hexmoor, H. (ed.), Proceedings of the International Conference on Integration of Knowledge Intensive Multi-Agent Systems. KIMAS’03, IEEE, 703–708. Gould, Stephen Jay; Vrba, Elisabeth 1982. Exaptation: A missing term in the science of form. Paleobiology 8: 4–15. Gray, Russell D. 1992. Death of the gene: Developmental systems fight back. In: Griffiths, Paul E. (ed.), Trees of Life: Essays in the Philosophy of Biology. Dordrecht: Kluwer, 165–209. Griesemer, James 2000. Development, culture, and the units of inheritance. Philosophy of Science 67 (Proceedings): S348–S368. Griffiths, Anthony J. F.; Gelbart, William M.; Miller, Jeffrey H.; Lewontin, Richard C. 1999. Modern Genetic Analysis. New York: W. H. Freeman & Co. Griffiths, Paul E. 2001. Genetic information: A metaphor in search of a theory. Philosophy of Science 68(3): 394–403. Griffiths, Paul E.; Neumann-Held, Eva 1999. The many faces of the gene. BioScience 49(8): 656–662. Grós, François 1989. Civilização do Gene. Lisboa: Terramar. Grosshans, Helge; Slack, Frank J. 2002. Micro-RNAs: small is plentiful. The Journal of Cell Biology 156(1): 17–21. Gudwin, Ricardo; Queiroz, João (eds.) 2007. Semiotics and Intelligent Systems Development. Hershey: IDEA GROUP. Guimarães, Romeu Cardoso 1986. O gene como uma molécula de RNA. Ciência e Cultura (Suppl.) 38: 939–940. — 1998. O conceito de informação em biologia. In: Encontro com as Ciências Cognitivas II, Gonzáles, Maria Eunice Q.; Broens, Mariana C. (orgs.). Marília: UNESP. Guimarães, Romeu Cardoso; Moreira, Carlos Henrique Costa 2000. O conceito sistêmico de gene — uma década depois. In: D’Ottaviano, Ítala M. L.; Gonzáles, Maria Eunice Q. (eds.), Auto-Organização: Estudos Interdisciplinares, vol. 2. (Coleção CLE [Centro de Lógica e Epistemologia] vol. 30.) Campinas: UNICAMP, 249–280. Gulick, Robert van 1993. Who is in charge here? And who’s doing all the work? In: Heil, John; Mele, Alfred R. (eds.), Mental Causation. Oxford: Oxford University Press, 233–256. Hall, Brian K. 2001. The gene is not dead, merely orphaned and seeking a home. Evolution and Development 3(4): 225–228. Hannon, Gregory J. 2002. RNA interference. Nature 418: 244–251. Hanson, Maureen R. 1996. Protein products of incompletely edited transcripts are detected in plant mitochondria. The Plant Cell 8(1): 1–3. Hastings, Kenneth E. M. 2005. SL trans-splicing: easy come or easy go? Trends in Genetics 21: 240–247. Hausman, Carl R. 1993. Charles Sanders Peirce’s Evolutionary Philosophy. Cambridge: Cambridge University Press.
References
235
Houser, Nathan 1997. Introduction: Peirce as a logician. In: Houser, Nathan; Roberts, Don D.; Evra, James van (eds.), Studies in the Logic of Charles Sanders Peirce. Bloomington: Indiana University Press, 1–22. Henikoff, Steven; Eghtedarzadeh, Mohammed K. 1987. Conserved arrangements of nested genes at the Drosophila Gart locus. Genetics 117: 711–725. Heylighen, Francis; Joslyn, Cliff 2001. Cybernetics and second order cybernetics. In: Meyers, Robert A. (ed.), Encyclopedia of Physical Science and Technology, vol. 4. New York: Academic Press, 155–170. Hieter, Philip; Boguski, Mark S. 1997. Functional genomics: It’s all how you read it. Science 278: 601–602. Hoffmeyer, Jesper 1996. Signs of Meaning in the Universe. Bloomington: Indiana University Press. — 1998. The unfolding semiosphere. In: Vijver, Gertrudis Van de; Salthe, Stanley; Delpos, Manuela (eds.), Evolutionary Systems: Biological and Epistemological Perspectives on Selection and Self-Organization. Dordrecht: Kluwer, 281–294. — 2002. The central dogma: A joke that became real. Semiotica 138(1/4): 1–13. — 2008a. Biosemiotics: An Examination into the Signs of Life and the Life of Signs. Scranton: University of Scranton Press. — 2008b. Introduction: Bateson the precurser. In: Hoffmeyer, Jesper (ed.), A Legacy for Living Systems: Gregory Bateson as Precursor to Biosemiotics. Dordrecht: Springer, 15–25. Hoffmeyer, Jesper; Emmeche, Claus 1991. Code-duality and the semiotics of nature. In: Anderson, Myrdene; Merrell, Floyd (eds.), On Semiotic Modeling. Berlin: Mouton de Gruyter, 117–166. [Reprinted as Hoffmeyer, Emmeche 2005.] Hoffmeyer, Jesper; Emmeche, Claus 2005. Code-duality and the semiotics of nature. Journal of Biosemiotics 1(1): 27–64. Hookway, Christopher 1985. Peirce. London: Routledge & Kegan Paul. — 1992. Peirce. London: Routledge & Kegan Paul. — 2002. Truth, Rationality, and Pragmatism: Themes from Peirce. Oxford: Oxford University Press. Hull, David L. 1974. The Philosophy of Biological Science. Englewood Cliffs: Prentice-Hall. Hulswit, Menno 2001. Semiotic and the cement of the universe: a Peircean process approach to causation. Transactions of the Charles S. Peirce Society 37(3): 339–363. — 2002. From Cause to Causation: A Peircean Perspective. Dordrecht: Kluwer. — 2005. How causal is downward causation? Journal for General Philosophy of Science 36(2): 1–26. Humphreys, Paul 1996. Aspects of emergence. Philosophical Topics 24(1): 53– 70. — 1997a. Emergence, not supervenience. Philosophy of Science 64 (proceedings): S337–S345. — 1997b. How properties emerge. Philosophy of Science 64(1): 1–17.
236
Genes, Information, and Semiosis
Ideker, Trey; Galitski, Timothy; Hood, Leroy 2001. A new approach to decoding life: Systems biology. Annual Review of Genomics and Human Genetics 2: 343–372. Jablonka, Eva 1994. Inheritance systems and the evolution of new levels of individuality. Journal of Theoretical Biology 170: 301–309. — 2001. The systems of inheritance. In: Oyama, Susan; Griffiths, Paul E.; Gray, Russell D. (eds.), Cycles of Contingency: Developmental Systems and Evolution. Cambridge: MIT Press, 99–116. — 2002. Information: Its interpretation, its inheritance, and its sharing. Philosophy of Science 69: 578–605. Jablonka, Eva; Lamb, Marion J. 2005. Evolution in Four Dimensions: Genetic, Epigenetic, Behavioral, and Symbolic Variation in the History of Life. Cambridge: MIT Press. Jablonka, Eva; Lamb, Marion J.; Avital, Eytan 1998. ‘Lamarckian’ mechanisms in Darwinian evolution. Trends in Ecology and Evolution 13: 206–210. Jablonka, Eva; Szathmáry, Eörs 1995. The evolution of information storage and heredity. Trends in Ecology and Evolution 10: 206–211. Jacob, François; Monod, Jacques 1961. On the regulation of gene activity. Cold Spring Harbour Symposium on Quantitative Biology 26: 193–211. Jakobson, Roman 1969. Linguistica e Comunicação. São Paulo: Cultrix. — 1973. Main Trends in the Science of Language. London: George Allen & Unwin. Johansen, Jorgen Dines 1993. Dialogic Semiosis. Bloomington: Indiana University Press. Joslyn, Cliff 2001. The semiotics of control and modeling relations in complex systems. BioSystems 60: 131–148. Judson, Horace Freeland 2001. Talking about the genome. Nature 409: 769. Kampa, Dione; Cheng, Jill; Kapranov, Philipp; Yamanaka, Mark; Brubaker, Shane; Cawley, Simon; Drenkow, Jorg; Piccolboni, Antonio; Bekiranov, Stefan; Helt, Gregg; Tammana, Hari and Gingeras, Thomas R. 2004. Novel RNAs identified from an in-depth analysis of the transcriptome of human chromosomes 21 and 22. Genome Research 14: 331–342. Kay, Lily E. 1995. Who wrote the book of life? Information and the transformation of molecular biology, 1945–1955. Science in Context 8: 609– 634. — 2000. Who Wrote the Book of Life? A History of the Genetic Code. Stanford: Stanford University Press. Keller, Evelyn Fox 2000. The Century of the Gene. Cambridge: Harvard University Press. — 2005. The century beyond the gene. Journal of Biosciences 30: 3–10. Kent, B. 1997. The interconnectedness of Peirce’s diagrammatic thought. In: Houser, Nathan; Roberts, Don D.; Evra, James van (eds.), Studies in the Logic of Charles Sanders Peirce. Bloomington: Indiana University Press, 445–459. Ketner, Kenneth Laine 1986. Peirce’s ‘most lucid and interesting paper’: An introduction to cenopythagoreanism. International Philosophical Quarterly 26: 375–392.
References
237
Kitano, Hiroaki 2002. Systems biology: a brief overview. Science 295: 1662– 1664. Kitcher, Philip 1982. Genes. British Journal for the Philosophy of Science 33: 337–359. Kim, Jaegwon 1993. Supervenience and Mind: Selected Philosophical Essays. Cambridge: Cambridge University Press. — 1996. Philosophy of Mind. Boulder: Westview Press — 1997. Supervenience, emergence, and realization in the philosophy of mind. In: Carrier, M.; Machamer, P. K. (eds.), Mindscape: Philosophy, Science, and the Mind. Pittsburgh: University of Pittsburgh Press, 271–293. — 1998. Mind in a Physical World: An Essay on the Mind-Body Problem and Mental Causation. Cambridge: The MIT Press. — 1999. Making sense of emergence. Philosophical Studies 95: 3–36. Klee, R. L. 1984. Micro-determinism and concepts of emergence. Philosophy of Science 51(1): 44–63. Klug, William S.; Cummings Michael R. 2002. Essentials of Genetics. (4th ed.) Upper Saddle River: Prentice Hall. Koller, B.; Fromm, H.; Galun, E.; Edelman, M.1987. Evidence for in vivo trans splicing of pre-mRNAs in tobacco chloroplasts. Cell 48: 111–119. Kornblihtt, A. R.; Pesce, C. G.; Alonso, C. R.; Cramer, P.; Srebrow, A.; Werbajh, S.; Muro, A. F. 1996. The fibronectin gene as a model for splicing and transcription studies. Faseb Journal 10: 248–257. Kornblihtt, A. R.; de la Mata, M.; Fededa, J. P.; Muñoz, M. J.; Nogues, G. 2004. Multiple links between transcription and splicing. RNA 10: 1489–1498. Kruse, F. 1994. Is Cosmic Evolution Semiosis? In: Moore, E.; Robin, R. (eds.), From Time and Chance to Consciousness: Studies in the Metaphysics of Charles Peirce. Oxford: Berg, 87–98. Kuhn, Thomas S. 1996 [1962]. The Structure of Scientific Revolutions (3rd ed.) Chicago: The University of Chicago Press. Kull, Kalevi 1999. Biosemiotics in the twentieth century: A view from biology. Semiotica 127(1/4): 385–414. Küppers, Bernd-Olaf 1990. Information and the origin of life. Cambridge: MIT Press. L = Peirce 1967. Lander, Eric S. et al. 2001. Initial sequencing and analysis of the human genome. Nature 409: 860–921. Langton, Christopher G. 1989. Artificial life. In: Langton, C. G. (ed.), Artificial Life. (Santa Fe Institute Studies in the Sciences of Complexity 6.) Redwood City: Addison-Wesley, 1–47. Lear, Jonathan 1988. Aristotle: The Desire to Understand. New York: Cambridge University Press. Lee, Tong Ihn; Nicola J. Rinaldi; François Robert; Duncam T. Odom; Ziv BarJoseph; Georg K. Gerber; Nancy M. Hannett; Christopher T. Harbison; Craig M. Thompson; Itamar Simon; Julia Zeitlinger; Ezra G. Jennings; Heather L. Murray; D. Benjamin Gordon; Bing Ren; John J. Wyrick; Jean-Bosco Tagne; Thomas L. Volkert; Ernest Fraenkel; David K. Gifford and Richard A. Young
238
Genes, Information, and Semiosis
2002. Transcriptional regulatory networks in Saccharomyces cerevisiae. Science 298: 799–804. Lenz, Guido 2005. The RNA interference revolution. Brazilian Journal of Medical and Biological Research 38(12): 1749–1757. Leite, Marcelo 2005. Biologia Total: Hegemonia e Informação no Genoma Humano. Campinas: Instituto de Filosofia e Ciências Humanas (IFCH), Universidade Estadual de Campinas (UNICAMP). Phd Thesis. Levins, Richard; Lewontin, Richard C. 1980. Dialectics and reductionism in ecology. Synthese 43: 47–78. Lewontin, Richard C. 1983. The organism as the subject and object of evolution. Scientia 118: 63–83. — 2000. The Triple Helix: Gene, Organism, and Environment. Cambridge: Harvard University Press. Lewin, Benjamin 2004. Genes VIII. Upper Saddle River: Pearson Education. Lizska, James J. 1990. Peirce’s interpretant. Transactions of the Charles S. Peirce Society 26(1): 17–61. Lodish, Harvey; Berk, Arnold; Matsudaira, Paul; Kaiser, Chris A.; Krieger, Monty; Scott, Matthew P.; Zipursky, S. Lawrence; Darnell, James 2003. Molecular Cell Biology. (5th ed.) New York: W. H. Freeman & Co. MacLennan, B. J. 2001. The emergence of communication through synthetic evolution. In: Patel, Mukesh; Honavar, Vasant; Balakrishnan, Karthik (eds.), Advances in the Evolutionary Synthesis of Intelligent Agents. Cambridge: MIT Press, 65–90. Magurran, Anne 2000. Backseat drivers, Review of The Century of the Gene by E. F. Keller. New York Times Book Reviews 10(December): 26. Maniatis, Tom; Reed, Robin 2002. An extensive network of coupling among gene expression machines. Nature 416: 499–506. Maniatis, Tom; Bosiljka Tasic 2002. Alternative pre-mRNA splicing and proteome expansion in metazoans. Nature 418: 236–243. Markoš, Anton 2002. Readers of the Book of Life: Contextualizing Developmental Evolutionary Biology. Oxford: Oxford University Press. Maull, Nancy 1977. Unifying science without reduction. Studies in History and Philosophy of Science 8: 143–162. Maynard Smith, John 1999. The Origins of Life: From the Birth of Life to the Origins of Language. Oxford: Oxford University Press. — 2000a. The concept of information in Biology. Philosophy of Science 67(2): 177–194. — 2000b. The Cheshire cat’s DNA. The New York Review of Books 47: 43–46. Maynard Smith, Jonh; Szathmáry, Eörs 1995. The Major Transitions in Evolution. Oxford: W. H. Freeman. — 1999. The Origins of Life: From the Birth of Life to the Origins of Language. Oxford: Oxford University Press. Mayr, Ernst 1982. The Growth of Biological Thought: Diversity, Evolution, and Inheritance. Cambridge: Harvard University Press. McCarthy, John 2004. Tackling the challenges of interdisciplinary bioscience. Nature Reviews, Molecular Cell Biology 5: 933–937.
References
239
Miyamoto, Shingo; Ben-Zion Katz; Robert M. Lafrenie and Kenneth M. Yamada 1998. Fibronectin and integrins in cell adhesion, signaling, and morphogenesis. Annals of the New York Academy of Sciences 857: 119–129. Merrell, Floyd 1995. Peirce’s Semiotics Now. Toronto: Canadian Scholar’s Press. Morgan, Conway Lloyd 1923. Emergent Evolution. London: Williams and Norgate. Modrek, Barmak; Lee, Christopher 2002. A genomic view of alternative splicing. Nature Genetics 30: 13–19. Morange, Michel 2006. Post-genomics, between reduction and emergence. Synthese 151(3): 355–360. Moss, Lenny 1992. A kernel of truth? On the reality of a genetic program. PSA 1: 335–348. — 2001. Deconstructing the gene and reconstructing molecular developmental systems. In: Oyama, Susan; Griffiths, Paul E.; Gray, Russell D. (eds.), Cycles of Contingency: Developmental Systems and Evolution. Cambridge: MIT Press, 85–97. — 2003. What Genes Can’t Do. Cambridge: MIT Press. Moyle, Leonie 2002. Most ingenious: troubles and triumphes of a century of genes. Biology and Philosophy 17: 715–727. MS = Peirce 1967. Murphey, Murray 1993. The Development of Peirce’s Philosophy. Cambridge: Harvard University Press. Nature 2005. In pursuit of systems (editorial). Nature 435: 1. NEM = Peirce 1976. Neumann-Held, Eva 1999. The gene is dead – Long live the gene: Conceptualizing genes the constructionist way. In Sociobiology and Bioeconomics: The Theory of Evolution in Biological and Economic Thinking, Peter Koslowski (Orgs.), 105–137. Berlin: Springer. — 2001. Let’s talk about genes: the process molecular gene concept and its context. In: Oyama, Susan; Griffiths, Paul E.; Gray, Russell D. (eds.), Cycles of Contingency: Developmental Systems and Evolution. Cambridge: MIT Press, 69–84. — 2006. Genes – causes – codes. Deciphering DNA’s ontological privilege. In: Neumann-Held, Eva M.; Rehmann-Sutter, Christoph (eds.), Genes in Development. Re-reading the Molecular Paradigm. Durham: Duke University Press, 238–271. Newman, D. V. 1996. Emergence and strange attractors. Philosophy of Science 63(2): 245–261. Nijhout, H. Frederik 1990. Metaphors and the role of genes in development. Bioessays 12(9): 441–446. Nilsen, Timothy W. 2001. Evolutionary origin of SL-addition trans-splicing: still an enigma. Trends in Genetics 17: 677–680. Nöth, Winfried (ed.) 1994. Origins of Semiosis: Sign Evolution in Nature and Culture. Berlin: Mouton de Gruyter. Nöth, Winfried 1995. Handbook of Semiotics. Bloomington: Indiana University Press.
240
Genes, Information, and Semiosis
— 2001. Protosemiotics and physicosemiosis. Sign Systems Studies 29(1): 13–26. O’Connor, Timothy 1994. Emergent properties. American Philosophical Quarterly 31(2): 91–104. Oyama, Susan 2000 [1985]. The Ontogeny of Information: Developmental Systems and Evolution. 2nd ed. Cambridge: Cambridge University Press. Oyama, Susan; Griffiths, Paul E.; Gray, Russell D. (eds.) 2001. Cycles of Contingency: Developmental Systems and Evolution. Cambridge: MIT Press. Pape, Helmuth 2002. What thought is for: the problematic identity of mental processes with chance events in Peirce’s idealistic metaphysics. Transactions of the Charles S. Peirce Society 38(1/2): 215–251. Pardini, Maria Inês; Campos, Moura; Guimarães, Romeu Cardoso 1992. A systemic concept of the gene. Revista Brasileira de Genética 15(3): 713–721. Parker, Kelly 1998. The Continuity of Peirce’s Thought. Nashville: Vanderbilt University Press. Pattee, Howard 1969. How does a molecule become a message? Developmental Biology Suppl. 3: 1–16. — 2000. Causation, control, and the evolution of complexity. In: Andersen et al. 2000: 63–77. Pearson, H. 2006. What is a gene? Nature 441: 399–401. Peirce, Charles Sanders 1931–1935. The Collected Papers of Charles Sanders Peirce. Electronic edition reproducing vols. I–VI [Hartshorne, C.; Weiss, P. (eds.), Cambridge: Harvard University Press, 1931–1935], vols. VII–VIII [Burks, A. W. (ed.), Cambridge: Harvard University Press, 1958]. Charlottesville: Intelex Corporation. [Here referred as CP, followed by volume and paragraph number.] — 1967. Annotated Catalogue of the papers of Charles S. Peirce. (Robin, R., ed.) Amherst: University of Massachusetts. [References to manuscripts and letters by Charles S. Peirce — MS and L — are in accordance with this catalogue.] — 1976. New Elements of Mathematics by Charles S. Peirce. (Eisele, Carolyn ed.) The Hague: Mouton. [Quoted as NEM, followed by page number]. — 1982–2000 [1839–1914]. Writings of Charles S. Peirce: a Chronological Edition, Vol. 2. Peirce Edition Project (ed.). Bloomington: Indiana University. [Quoted as W, followed by page number] — 1998 [1893–1913]. The Essential Peirce: Selected Philosophical Writings. Vol. II. (Peirce Edition Project, ed.) Bloomington and Indianapolis: Indiana University Press. [Here referred as EP2, followed by the number of the page.] Peltonen, Lenna; McKusick, Victor A. 2001. Dissecting human disease in the postgenomic era. Science 291: 1224–1229. Pennisi, Elizabeth 2001. Behind the scenes of gene expression. Science 293: 1064–1067. — 2003. Tracing life’s circuitry. Science 302: 1646–1649. Pessoa jr., Osvaldo 1996. Medidas Sistêmicas e Organização. In: Debrun, Michel; Gonzales, Maria Eunice; Pessoa jr., Osvaldo (eds.), Auto-Organização. Campinas: CLE (18), 129–161.
References
241
Petrilli, Susan; Ponzio, Augusto 2005. Semiotics Unbounded: Interpretive Routes through the Open Network of Signs. Toronto: University of Toronto Press. Pietarinen, Ahti-Veikko 2005. Signs of Logic: Peircean Themes in the Philosophy of Language, Games, and Communication. Berlin: Springer. Pihlström, Sami 2002. The re-emergence of the emergence debate. Principia 6: 133–181. Polanyi, Michael 1968. Life’s irreducible structure. Science 160: 1308–1312. Popper, Karl R.; Eccles, John C. 1986 [1977]. The Self and Its Brain. London: Routledge and Kegan Paul. Portin, Petter 1993. The concept of the gene: short history and present status. Quarterly Review of Biology 68: 173–223. Potter, Vincent G. 1997 [1967]. Charles S. Peirce: On Norms & Ideals. New York: Fordham University Press [The University of Massachusetts Press]. Queiroz, João 2003. Comunicação simbólica em primatas não-humanos: uma análise baseada na semiótica de C.S.Peirce. Rev Bras Psiquiatr 25(Supl II): 2–5. Queiroz, João; El-Hani, Charbel Niño 2004. Towards a multi-level approach to the emergence of semiosis in semiotic systems. Technical Reports DCAFEEC-UNICAMP, DCA07-04, 1–21. — 2006a. Semiosis as an emergent process. Transactions of the Charles S. Peirce Society 42(1): 78–116. — 2006b. Towards a multi-level approach to the emergence of meaning processes in living systems. Acta Biotheoretica 54(3): 174–206. Queiroz, João; Emmeche, Claus; El-Hani, Charbel Niño 2005. Information and semiosis in living systems: A biosemiotic approach. S.E.E.D. Journal: Semiotics, Evolution, Energy, and Development 5(1): 60–90. Queiroz, João; Emmeche, Claus; El-Hani, Charbel Niño 2008. A Peircean approach to information and its relationship with Bateson’s and Jablonka’s Ideas. The American Journal of Semiotics 24(1/3): 75–94. Queiroz, João; Merrell, Floyd (guest eds.) 2005. Abduction — between subjectivity and objectivity. Semiotica [Special Issue on Abductive Inference] 153(1/4). Queiroz, João; Merrell, Floyd 2006. Semiosis and pragmatism: Toward a dynamic concept of meaning. Sign Systems Studies 34(1): 37–66. Queiroz, João; Ribeiro, S. 2002. The biological substrate of icons, indexes and symbols in animal communication: a neurosemiotic analysis of Vervet monkey alarm-calls. In: Shapiro, M. (ed.), The Peirce Seminar Papers 5. New York: Berghahn Books, 69–78. Ransdell, Joseph 1977. Some leadings ideas of Peirce’s semiotic. Semiotica 19(3/4): 157–178. Rescher, Nicholas 1996. Process Metaphysics: An Introduction to Process Philosophy. New York: SUNY Press.
242
Genes, Information, and Semiosis
— 2002. Process philosophy. In: Zalta, E. N. (ed.). The Stanford Encyclopedia of Philosophy (Summer 2002 Edition).121 Rheinberger, Hans-Jörg 2000. Gene concepts: fragments from the perspective of molecular biology. In: Beurton, Peter; Falk, Raphael; Rheinberger, Hans-Jörg (eds.), The Concept of the Gene in Development and Evolution. Cambridge: Cambridge University Press, 219–239. Ribeiro, S.; Loula, A.; Araújo, I.; Gudwin, R.; Queiroz, J. 2007. Symbols are not uniquely human. Biosystems 90: 263–272. Rios, Ricardo Iglesias 2004. O início do fim do gene, resenha de Keller, E. F. O Século do Gene. Ciência Hoje 34(204): 72–73. Roederer, Juan G. 2005. Information and Its Role in Nature. Berlin: Springer. Ronald, E. M. A.; Sipper, M.; Capcarrère, M. S. 1999. Design, observation, surprise! A test for emergence. Artificial Life 5: 225–239. Ross, David 1995 [1923]. Aristotle. Routledge: London. Salthe, Stanley N. 1985. Evolving Hierarchical Systems: Their Structure and Representation. New York: Columbia University Press. Santaella-Braga, Lucia 1994. Peirce’s broad concept of mind. European Journal for Semiotic Studies 6(3/4): 399–411. Sarkar, Sahotra 1996. Biological information: A skeptical look at some central dogmas of molecular biology. In: Sarkar, Sahotra (ed.), The Philosophy and History of Molecular Biology: New Perspectives. Dordrecht: Kluwer, 187– 231. — 1998a. Genetics and Reductionism. Cambridge: Cambridge University Press. — 1998b. Forty years under the central dogma. Trends in Biochemical Sciences 23: 312–316. — 2000. Information in genetics and developmental biology: Comments on Maynard Smith. Philosophy of Science 67(2): 208–213. Sampedro, Javier 2004. Deconstruyendo a Darwin: Los Enigmas de la Evolución a la Luz de la Nueva Genética. Barcelona: Crítica. Savan, David 1986. Response to T. L. Short. Transactions of the Charles S. Peirce Society 22(2): 125–143. — 1988. An Introduction to C. S. Peirce’s Full System of Semeiotic. (Toronto Semiotic Circle Monographs 1.) Toronto: Toronto Semiotic Circle. Schaffner, Kenneth F. 1967. Approaches to reduction. Philosophy of Science 34: 137–147. Schneider, Tom D. 1994. Sequence logos, machine/channel capacity, Maxwell’s demon, and molecular computers: A review of the theory of molecular machines. Nanotechnology 5: 1–18. Schneider, Tom D.; Stephens, R. M. 1990. Sequence logos: A new way to display consensus sequences. Nucleic Acids Research 18: 6097–6100. Schröder, Jürgen 1998. Emergence: Non-deducibility or Downwards Causation? Philosophical Quarterly 48(193): 433–452.
121
Available at http://plato.stanford.edu/archives/ sum2002/entries/processphilosophy/ (accessed in July 11, 2005).
References
243
Schulz, R. A.; Cherbas, L.; Cherbas, P. 1986. Alternative splicing generates two distinct Eip28/29 gene transcripts in Drosophila Kc cells. Proceedings of the National Academy of Sciences of the United States 83, 9428–9432. Scott, Alwyn C. 2004. Reductionism revisited. Journal of Consciousness Studies 11(2): 51–68. Seibt, Johanna 1996. The myth of substance and the fallacy of misplaced concreteness. Acta Analytica 15: 61–76. Shannon, Claude E. 1948. A mathematical theory of communication. Bell System Technical Journal 27: 379–423, 623–656. Shannon, Claude E.; Weaver, Warren 1949. The Mathematical Theory of Communication. Urbana: University of Illinois Press. Shin, Sun-Joo 2002. The Iconic Logic of Peirce’s Graphs. Cambridge: MIT Press. Short, Thomas 1996. Interpreting Peirce’s interpretant: a response to Lalor, Liszka, and Meyers. Transactions of the Charles S. Peirce Society: A Quarterly Journal in American Philosophy 32(4): 488–541. — 1998. What’s the use? Semiotica 122(1/2): 1–68. — 2002. Darwin’s concept of final cause: neither new nor trivial. Biology and Philosophy 17: 323–340. — 2007. Peirce’s Theory of Signs. Cambridge: Cambridge University Press. Smith, Kelly C. 1994. The Emperor’s New Genes: the Role of the Genome in Development and Evolution. Unpublished PhD thesis. Durham: Duke University. Skagestad, Peter 2004. Peirce’s Semeiotic Model of the Mind. In: The Cambridge Companion to Peirce, Cheryl Misak (ed.), Cambridge, UK: Cambridge University Press, 241–256. Sorek, Rotem; Shamir, Ron; Ast, Gil 2004. How prevalent is functional alternative splicing in the human genome? Trends in Genetics 20(2): 68–71. Sperry, R. W. 1991. In defense of mentalism and emergent interaction. Journal of Mind and Behavior 12: 221–245. Stent, Gunther S. 1968. That was the molecular biology that was. Science 160: 390–395. Stephanopoulos, Gregory; Alper, Hal; Moxley, Joel 2004. Exploiting biological complexity for strain improvement through systems biology. Nature Biotechnology 22: 1261–1267. Stephan, Achim 1997. Armchair arguments against emergentism. Erkenntnis 46: 305–314. — 1998. Varieties of emergence in artificial and natural systems. Zeitschrift für Naturforschung C (A Journal of Biosciences): 53(7/8): 639–656. — 1999. Emergenz: Von der Unvorhersagbarkeit zur Selbstorganisation. Dresden: Dresden University Press. Stephens, Craig 1998. Bacterial sporulation: A question of commitment? Current Biology 8: R45–R48. Sterelny, Kim 2000. The ‘genetic program’ program: A commentary on Maynard Smith on information in Biology. Philosophy of Science 67(2): 195–201. — 2001. Dawkins vs. Gould: Survival of the Fittest. Cambridge: Icon Books.
244
Genes, Information, and Semiosis
Sterelny, Kim; Griffiths, Paul E. 1999. Sex and Death: An Introduction to Philosophy of Biology. Chicago: The University of Chicago Press Stjernfelt, Frederik 1992. Categorical perception as a general prerequisite to the formation of signs? On the biological range of a deep semiotic problem in Hjelmslev’s as well as Peirce’s semiotics. In: Sebeok, Thomas A.; UmikerSebeok, Jean (eds.), Biosemiotics: The Semiotic Web 1991. Berlin: Mouton de Gruyter, 427–454. — 2006. The semiotic body: A semiotic concept of embodiment? In: Nöth, Winfried (ed.), Semiotic Bodies, Aesthetic Embodiment, and Cyberbodies. Kassel: Kassel University Press, 13–48. [Also as Ch. 12 in Stjernfelt 2007.] — 2007. Diagrammatology: An Investigation on the Borderlines of Phenomenology, Ontology, and Semiotics. Dordrecht: Springer. Stotz, Karola; Griffiths, Paul E.; Knight, Rob 2004. How biologists conceptualize genes: An empirical study. Studies in the History and Philosophy of Biological & Biomedical Sciences 35: 647–673. Strohman, Richard 2002. Maneuvering in the complex path from genotype to phenotype. Science 296: 701–703. Stuart, C. I. J. M. 1985. Bio-informational equivalence. Journal of Theoretical Biology 113: 611–636. Symons, John 2002. Emergence and reflexive downward causation. Principia, Special Issue on Emergence and Downward Causation (El-Hani, C. N. Ed.) 6: 183–202. Szathmáry, Eörs; Jordán, Ferenc; Pál, Csaba 2001. Can genes explain biological complexity? Science 292: 1315–1316. Thompson, J. N.; Pellmyr, O. 1991. Evolution of oviposition behavior and host preference in Lepidoptera. Annual Review of Entomology 36: 65–89. Tienne, André De 1992. Peirce’s semiotic monism. In: Deledalle, Gérard; Balat, Michel; Deledalle Rhodes, Janice (eds.), Signs of Humanity / L ’homme et ses signes: Proceedings of the IVth International Congress / Actes du IVe Congrès Mondial. 3 vols. (Approaches to Semiotics 107.) Berlin: Mouton de Gruyter, 1291–1303. — 2003. Learning qua semiosis. S.E.E.D. Journal (Semiotics, Evolution, Energy, and Development) 3(3): 37–53. Tiercelin, Claudine 1995. The relevance of Peirce’s semiotic for contemporary issues in cognitive science. In L.Haaparanta & S.Heinamaa (eds.), Mind and Cognition: Philosophical Perspectives on Cognitive Science and Artificial Intelligence. (= Acta Philosophica Fennica 58: 37–74). Tupler, Rossella; Perini, Giovanni; Green, Michael 2001. Expressing the human genome. Nature 409: 832–833. Vandenberghe, Amanda E.; Thomas H. Meedel and Kenneth E.M. Hastings 2001. mRNA 5’-leader trans-splicing in the chordates. Genes & Development 15: 294–303. Venter, Craig et al. 2001. The sequence of the human genome. Science 291: 1305–1351.
References
245
Vijver, Gertrudis Van de; Salthe, Stanley N.; Delpos, M. (eds.) 1998. Evolutionary Systems: Biological and epistemological perspectives on selection and self-organization. Dordrecht: Kluwer. Violi, Patrizia 1999. Semiotics and cognition. In: Wilson, Robert A.; Keil, Frank C. (eds.), The MIT Encyclopedia of the Cognitive Sciences, Cambridge: MIT Press, 744–745. W = Peirce 1982–2000. Waal, Cornelis de 2001. On Peirce. Wadsworth: Thomson Learning. Wang, Wen; Zhang, Jianming; Alvarez, Carlos; Llopart, Ana; Long, Manyuan 2000. The origin of the Jingwei gene and the complex modular structure of its parental gene, Yellow Emperor, in Drosophila melanogaster. Molecular Biology and Evolution 17(9): 1294–1301. Wanscher, Johan Henrik 1975. The history of Wilhelm Johannsen’s genetical terms and concepts from the period 1903 to 1926. Centaurus 19(2): 125–147. Waters, C. Kenneth 1994. Genes made molecular. Philosophy of Science 61: 163– 185. — 2000. Molecules made biological. Revue Internationale de Philosophie 54: 539–564. Watson, James D. 2000. A Passion for DNA. Cold Spring Harbor: CSHL Press. Watson, James D.; Berry, Andrew 2003. DNA: The Secret of Life. New York: Alfred Knopf. Watson, James D.; Crick, Francis H. C. 1953. A structure for deoxyribose nucleic acid. Nature 171: 737–738. Wilden, Anthony 1980. System and Structure. Essays in Communication and Exchange. 2nd rev. ed. [1.ed. 1972]. New York: Tavistock Publications. Wilkins, Adams S. 2002. Grappling with developmental complexity. BioEssays 24: 1193–1195. Williams, George C. 1966. Adaptation and Natural Selection. Princeton: Princeton University Press. Williams, Nigel 1997. Biologists cut reductionist approach down to size. Science 277(5325): 476–477. Wimsatt, William C. 1976. Reductive explanation: a functional account. In: Cohen, R. S. (ed.), PSA 1974. Dordrecht: Reidel, 671–710. Wright, Fred; Lemon, William J.; Zhao, Wei D.; Sears, Russell; Zhuo, Degen; Wang, Jian-Ping; Yang, Hee-Yung; Baer, Troy; Stredney, Don; Spitzner, Joe; Stutz, Al; Krahe, Ralf; Yuan, Bo 2001. A draft annotation and overview of the human genome. Genome Biology 2: 1–18. Wright, Larry 1973. Functions. Philosophical Review 82: 139–168 (also printed in Colin Allen, Marc Bekoff & Georg Lauder (eds.) 1998. Nature’s Purposes. Analyses of function and design in biology. Cambridge: MIT Press.) Wynnie, J. A. 2000. Information and structure in molecular biology: Comments on Maynard Smith. Philosophy of Science 67(3): 517–526. Yockey, Hubert P. 1992. Information Theory and Molecular Biology. Cambridge: Cambridge University Press.
Index A Adami, 77, 79, 100, 217 Adams, 78 Alberts, 115 Alexander, 164, 168, 215 Andersen, 146, 155 Aristotle, 160 Arnellos, 198, 208 Ashburner, 69 Ast, 216 Azzone, 146 Atlan, 17, 221 Avital, 77, 83
B Baas, 146 Bailey, 171 Barabási, 53, 81 Barbieri, 19, 124 Bateson, 99–103, 106, 109, 110, 111 Beadle, 35 Beckermann, 146 Bedau, 146, 147 Benzer, 31, 49 Bergandi, 53 Bergman, 93, 94 Berry, 54 Bertalanffy, 163 Black, 41, 42, 119 Blitz, 146, 166, 167 Boguski, 54 Boogerd, 153, 154, 175, 176 Brenner, 19 Brier, 102 Britten, 38 Broad, 153 Brosius, 69 Brown, 69 Bruggeman, 225 Bruni, 18, 53, 84, 100, 198, 208 Brunning, 87–89, 171
Bunge, 163 Burch, 89 Burian, 22, 33, 58
C Camargo, 54 Campbell, 117, 140, 161 Cangelosi, 146 Cariani, 146, 147, 170 Carlson, 29, 58 Caro, 149 Carroll, 49 Cavalier-Smith, 216 Chong, 16, 53, 54 Clark, 83 Claverie, 52, 53, 216, 217 Colapietro, 87 Cover, 78 Coyne, 20 Crick, 16, 29, 32, 33, 35 Csete, 16 Cummings, 115 Cummins, 104, 105 Cunningham, 146
D Danen, 42 Dawkins, 62 Deacon, 87, 165 Debrock, 92 Doyle, 16
E Eccles, 150 El-Hani, 17, 65, 135, 139, 142, 146, 147, 152, 155, 156, 159, 160, 161, 166, 198, 208, 218 Emmeche, 18, 19, 21, 65, 77, 80, 90, 100, 101, 134, 146, 148, 152, 155–
247
Index 158, 165, 169, 176, 185, 186, 208, 218 Enard, 47 Epp, 17, 41, 43, 60, 63
H
Falk, 15, 19, 22, 27–31, 33, 37, 38, 51, 57–60, 62, 68, 72 Favareau, 19 Fetzer, 163, 166 Fire, 47 Fitzgerald, 94 Fogle, 19, 29, 34, 37, 39–41, 43–46, 51, 52, 58–61, 65, 67, 68, 70–73, 134, 214 Forster, 87 Freadman, 87 Freeman, 87 Futuyma, 215
Hall, 20, 22, 58, 198, 201, 224, 225 Hannon, 47 Hanson, 44 Harries-Jones, 103 Hartshorne, 164, 215 Hastings, 46 Hausman, 178 Henikoff, 44 Heylighen, 102 Hieter, 54 Hoffmeyer, 18, 19, 21, 77, 80, 87, 100, 101, 110, 134, 208 Hood, 18 Hookway, 87, 88, 172 Houser, 87, 88, 171, 172 Hull, 32 Hulswit, 155–157, 159–161 Humphreys, 146
G
I
F
Gabius, 83 Galef Jr., 83 Galitski, 18 Gavin, 133, 134 Gelbart, 15, 20, 135 Gelder, 163 Gilbert, 218 Gillett, 146 Godfrey-Smith, 77, 105, 126 Goldschmidt, 30, 40 Gomes, 142 Gould, 17, 69, 70 Gray, 22 Griesemer, 83 Griffiths, 17–19, 30, 45, 58, 61–64, 77– 80, 126, 214 Griffiths A., 115, 123 Grós, 15 Grosshans, 47 Gudwin, 87 Guimarães, 17, 19, 37, 42, 43, 45, 47, 52, 70–73, 77, 121, 122, 134, 135, 214, 215, 224
Ideker, 16, 18, 52, 81, 220, 223
J Jablonka, 77–79, 82–84, 90, 99, 100, 103–111, 183, 209, 211, 224 Jacob, 36, 37, 49 Jakobson, 87 Johannsen, 15, 27, 28 Johansen, 89, 92 Joslyn, 102, 170 Judson, 20, 66
K Kampa, 57 Karpiuk, 83 Kay, 34, 77 Keller, 15–22, 29, 32, 36, 37, 38, 41– 43, 45, 47, 49, 51–55, 58, 59, 66, 219, 220, 223, 224 Kent, 172
I
Genes, Information, and Semiosis
248
Ketner, 89 Kim, 145, 146, 156, 161, 162 Kitano, 16, 53 Kitcher, 22, 29, 31–33, 58, 61, 63 Klee, 146 Klug, 115 Kohne, 38 Koller, 46 Koppel, 17, 221 Kornblihtt, 41, 42, 134 Kruse, 147 Kuhn, 20, 56, 115 Kull, 19 Küppers, 99, 100
L Lamb, 77, 83 Lander, 47, 48, 57, 216 Langton, 146 Lear, 160 Lee, 48, 49 Leite, 34, 47, 48, 53, 54, 58 Lenz, 47 Lewin, 44, 72, 115–117, 121, 122 Levins, 53 Lewontin, 18, 53 Liszka, 91 Lloyd Morgan, 167 Lodish, 115, 118
M MacArthur, 149 MacLennan, 146 Magen, 216 Magurran, 20 Maniatis, 49, 68, 134, 217 Markoš, 19 Maull, 32 Maynard Smith, 20, 77, 80, 95, 100, 105, 133 Mayr, 29, 30, 149 McCarthy, 53 McKusick, 49, 56 Mendel, 27, 28
Merrell, 87–89, 142 Miyamoto, 42 Modrek, 48 Monod, 36, 37, 49 Morange, 53 Moreira, 17, 70–72, 121, 122, 135, 215 Morgan, 28, 29, 148, 164, 166–169, 215 Morris, 89 Moss, 15, 17, 22, 58, 65, 66, 68, 127, 128, 215, 217–220 Moyle, 20, 52 Muller, 29, 30, 32, 40 Murphey, 88, 166, 177
N Neumann-Held, 17, 19, 22, 30, 45, 58, 61–64, 73, 79, 80, 214, 215, 218, 219 Newman, 146 Nijhout, 17, 224 Nilsen, 46 Nurse, 81 Nöth, 89, 110, 129, 147
O O’Connor, 146 Oltvai, 53, 81 Oyama, 17–19, 77, 79, 184, 223
P Pape, 178 Pardini, 19, 37, 43, 45, 47, 52, 70–73, 134, 214, 224 Parker, 88, 147, 166, 172, 175 Pattee, 66, 170 Pearson, 19 Peirce, 15, 88, 89, 91–95, 101, 103, 105, 107–109, 123–125, 132, 135, 139, 141, 144, 147, 150, 164–166, 169, 170–172, 175, 179, 201, 215, 222, 223
249
Index Pellmyr, 83 Peltonen, 49, 56 Pennisi, 16, 65 Pereira, 146, 159 Pessoa, 163 Petrilli, 87 Pietarinen, 87 Pihlström, 146, 159 Polanyi, 140, 156 Ponzio, 87 Popper, 150 Portin, 19, 20, 29 Potter, 88, 93, 144, 166, 175, 177
Q Queiroz, 65, 83, 87, 89, 99, 135, 139, 142, 152, 160, 161, 166, 198, 208
R Ransdell, 89, 164, 165 Ray, 16, 53, 54 Reece, 117 Reed, 134 Rescher, 89, 96, 164, 209, 215 Rheinberger, 60, 68, 214, 224 Ribeiro, 83 Rios, 20 Roederer, 100 Ronald, 146 Ross, 160
S Salthe, 139, 140, 142, 144, 145, 178 Sampedro, 134 Santaella-Braga, 89 Sarkar, 16, 17, 19, 32, 34, 77, 78, 79, 218 Savan, 88, 91, 142, 171 Savigny, 146 Schaffner, 32 Schneider, 99 Schröder, 146, 155
Schulz, 44 Scott, 158 Seibt, 164, 215 Shannon, 78, 99, 100 Shin, 87 Short, 87, 91, 132, 133 Skagestad, 87 Slack, 47 Smith, 17 Sorek, 48 Szathmáry, 49, 77, 95, 217 Stent, 36 Stephan, 146–152, 155, 162 Stephanopoulos, 16, 53 Stephens, 54, 99 Sterelny, 62, 77 Stjernfelt, 19, 87, 110, 121 Stotz, 16, 58, 60 Strohman, 56 Stuart, 19, 77–79 Sturtevant, 30 Symons, 146, 156
T Taborsky, 103 Tasic, 49, 68, 217 Tatum, 35 Thomas, 78 Thompson, 83 Tienne, 93, 141, 171 Tiercelin, 87 Tupler, 55, 56 Turner, 146
U Ushenko, 164, 215
W Waal, 172 Vandenberghe, 46 Wang, 45, 56, 57 Wanscher, 27
250
Genes, Information, and Semiosis
Waters, 22, 43, 52, 60, 61, 67 Watson, 16, 29, 32, 33, 54 Weaver, 78, 99, 100 Weiss, 164, 215 Venter, 43, 47, 48, 53, 216 Vest, 55 Whitehead, 164, 215 Whitelaw, 65 Videira, 146, 156, 159 Vijver, 159 Wilden, 101 Wilkins, 20 Williams, 62, 81, 223
Wimsatt, 32 Violi, 87 Vrba, 70 Vries, 15 Wright, 48, 104, 105 Wynnie, 77
Y Yamada, 42 Yockey, 99, 100
251
The authors Charbel Niño EL-HANI is a Professor at the Institute of Biology, Federal University of Bahia, Brazil . His research interests are philosophy of biology, biosemiotics, science education research, and animal behavior. His publications include ‘Gene concepts in higher education cell and molecular biology textbooks’ (with M. A. Pitombo and A. M. R. Almeida, 2008, Science Education International 19); ‘Multicultural education, pragmatism, and the goals of science teaching’ (with E. F. Mortimer, 2007, Cultural Studies of Science Education 2); ‘Between the cross and the sword: the crisis of the gene concept’ (2007, Genetics and Molecular Biology 30); ‘Towards a multi-level approach to the emergence of meaning processes in living systems’ (with J. Queiroz, 2006, Acta Biotheoretica 54); ‘A semiotic analysis of the genetic information system’ (with J. Queiroz and C. Emmeche, 2006, Semiotica 160); ‘Semiosis as an emergent process’ (with J. Queiroz, 2006, Transactions of the Charles Sanders Peirce Society 42); ‘Information and semiosis in living systems: A biosemiotic approach’ (with J. Queiroz and C. Emmeche, 2005, S.E.E.D. Journal 5); ‘On some theoretical grounds for an organism-centered biology: Property emergence, supervenience, and downward causation’ (with C. Emmeche, 2000, Theory in Biosciences 119). Evolução: O Sentido da Biologia (with D. Meyer, UNESP Publishing House, Brazil). João QUEIROZ is a Professor at the Graduate Studies Program on History, Philosophy, and Science Teaching (UFBA/UEFS), Federal University of Bahia, Brazil, director of the Group for Research on Artificial Cognition (UEFS) and associate researcher at the Dept. of Computer Engineering and Industrial Automation, State University of Campinas . His research interests are Peirce’s semiotics, american pragmatism and cognitive science. His publications include Computação, Cognição, Semiose (with A. Loula and R. Gudwin, 2007, EDUFBA; Semiotics and Intelligent Systems Development (with R. Gudwin, 2007, Idea Group); Advanced Issues in Cognitive Science and Semiotics (with P. Farias, 2006, Shaker Verlag); Semiose segundo Peirce (2004, EDUC); ‘Abduction — between subjectivity and objectivity’ (with F. Merrell, 2005, Semiotica 153).
252
Genes, Information, and Semiosis
Claus EMMECHE is Associate Professor and Director at the Center for the Philosophy of Nature and Science Studies, University of Copenhagen, Denmark . His research interests are biosemiotics, philosophy of biology, and theoretical biology. His publications include ‘On some theoretical grounds for an organism-centered biology: Property emergence, supervenience, and downward causation’ (with C. N. El-Hani, 2000, in Theory in Biosciences 119); Reading Hoffmeyer, Rethinking Biology (with K. Kull and Frederik Stjernfelt, 2002, Tartu Semiotics Library 3), ‘The chicken and the Orphean egg: On the function of meaning and the meaning of function’ (2002, in Sign Systems Studies 30); ‘Causal processes, semiosis, and consciousness’ (2004, in Johanna Seibt (ed.), Process Theories: Crossdisciplinary Studies in Dynamic Categories. Dordrecht: Kluwer).
Tartu Semiotics Library Series editors: Peeter Torop, Kalevi Kull, Silvi Salupere Vol. 1, 1998
V. V. Ivanov, J. M. Lotman, A. M. Pjatigorski, V. N. Toporov, B. A. Uspenskij Theses on the Semiotic Study of Cultures
Vol. 2, 1999
J. Levchenko (ed.) Conceptual Dictionary of the Tartu-Moscow Semiotic School
Vol. 3, 2002
C. Emmeche, K. Kull, F. Stjernfelt Reading Hoffmeyer, Rethinking Biology
Vol. 4, 2005
J. Deely Basics of Semiotics (4th edition, bilingual)
Vol. 5, 2006
M. Grishakova The Models of Space, Time and Vision in V. Nabokov's Fiction: Narrative Strategies and Cultural Frames
Vol. 6, 2008
P. Lepik Universaalidest Juri Lotmani semiootika kontekstis
Vol. 7, 2008
P. Lepik Universals in the Context of Juri Lotman’s Semiotics
Vol. 8, 2009
C. N. El-Hani, J. Queiroz, C. Emmeche Genes, Information, and Semiosis