174 26 18MB
English Pages 180 [192] Year 1977
Linguistische Arbeiten
54
Herausgegeben von Herbert E. Brekle, Hans Jürgen Heringer, Christian Rohrer, Heinz Vater und Otmar Werner
Stefan R. Fink
Aspects of a pedagogical grammar based on case grammar and valence theory
Max Niemeyer Verlag Tübingen 1977
CIP-Kurztitelaufnahme der Deutschen Bibliothek Fink, Stefan R. Aspects of pedagogical grammar : based on case grammar and valence theory. 1. Aufl. Tübingen : Niemeyer, 1977. (Lingusitische Arbeiten ; 54) ISBN 3^84- 10287-X
ISBN 3-484-10287-x © Max Niemeyer Verlag Tübingen 1977 Alle Rechte vorbehalten. Ohne ausdrückliche Genehmigung des Verlages ist es auch nicht gestattet, dieses Buch oder Teile daraus auf photomechanischem Wege zu vervielfältigen. Printed in Germany
TABLE OF CONTENTS
Page ACKNOWLEDGEMENTS 1.
2.
ix
SOME REFLECTIONS ON THE RELATIONSHIP BETWEEN LINGUISTIC THEORIES AND PEDAGOGICAL GRAMMARS
1
1.1 On the concept of 'pedagogical granmar' 1.2 Which linguistic theory?
1 2
1.3 The universality of a pedagogical granmar
4
THEORETICAL BACKGROUND
6
2.1
6 6 7
Linguistic valence—a first approximation 2.1.1 Valence—'a metaphorical terminology1 2.1.2 Tesniere's concept of valence 2.1.3 The valence concept (Valenzbegriff) in German linguistics 2.1.3.1 Different interpretations of valence
...
2.1.3.2 Helbig's concept of valence 2.1.3.3 Engel's 'sentence models' (Satzbaupläne)— Valence research at the Institut für Deutsche Sprache 2.1.3.4 Vater's revised generative dependency model 2.2
Thoughts on the isolation of norpho-syntactic units and their relation to semantic concepts; norpho-syntactic valence versus logico-semantic valence 2.2.1 Semantics and syntax 2.2.2 Isolating syntactic constituents 2.2.3 Syntactic constituents as a testing ground for the nuclear/peripheral (obligatory/free distinction
10 10 13
2O 24
26 26 27
32
VI
2.2.3.1
Regarding the distinction obligatory/ facultative1 versus 'free 1 2.2.4 Towards a logico-semantic specification of valence 2.2.4.1 Case graimiar in the context of linguistic valence 2.2.4.1.1 Criteria for case identification 2.2.4.2
3.
2.2.4.1.2 Nuclear versus peripheral cases . Semantic sentence structures as representations of cognitive concepts
33 38 39 4O 41 42
2.2.4.3 Features as criteria for the identification of relational constants 2.2.4.4 The non-causative relational constants . . . 2.2.4.4.1 The Object relation 2.2.4.4.2 The Experiencer relation . . . .
50 5O 53
2 . 2 . 4 . 4 . 3 The Benefactive relation
55
. . . .
2 . 2 . 4 . 4 . 4 The Locative relation
57
2.2.4.4.5 The Time relation , 2.2.1.5 Summary of relational constants selected for consideration in a pedagogical granmar
58
CONTRIBUTIONS TO A PEDAGOGICAL GRAMMAR—THE APPLICATION OF THEORETICAL CONSIDERATIONS
62
3.1 The experimental language test in German as a research tool in the investigation of errors in the realization of logico-semantic valence 3.1.1 Objective of the test
62 62
3.1.2 Test design 3.1.3 Test administration 3.1.4 Selection of verbs (predicates) for valence analysis 3.2 Discussion of selected verbs based on presented theoretical considerations—possible application in a pedagogical grammar 3.2.1 On the Agent-Object-Experiencer relation 3.2.1.1 antworten/frag en 'to answer'/'to ask1 3.2.1.2 Sprechen/sagen/reden/erzählen; 'to speak'/'to say'/'to talk'/'to tell, narrate1
59
...
63 64 64
67 68 68
75
vii
3.2.1.3
grüßen/begvüßen/grüßen lassen/ empfangen 'to greet"/'to give regards to'/'to welcome1
3.2.2 On the Benefactive notion 3.2.2.1 State/Benefactive 3.2.2.2 Process/Benefactive 3.2.2.3 Action/Benefactive 3.2.2.4 Action-Process/Benefactive 3.2.3 Application of logico-semantic valence to locative verb types 3.2.3.1 On the State/Locative relation 3.2.3.2 Action-Process/Locative 3.2.3.3 Action/Locative and Process/Locative . . . . 3.2.4 Observations on reflexivity 3.2.4.1 Various types of reflexivity
3.2.4.2
3.2.4.1.1 Facultative reflexivity 3.2.4.1.2 Obligatory reflexivity 3.2.4.1.3 Semireflexivity Special functions of reflexivity
3.2.4.2.1 Proform for passive 3.2.4.2.2 Stative passive versus state reflexive 3.2.4.2.3 State reflexive/stative passive and attributive function 3.2.4.3 Test items with the reflexivity function 3.2.4.3.1 Ausziehen 'to take off clothes', 'undress' 3.2.4.3.2 Ansehen 'to look at1 3.2.4.3.3 Ändern 'to change1 3.2.4.3.4 Further problems with reflexivity 3.2.5 On reciprocal verbs 3.2.6 Verbs implying time relations 3.2.7 On predicates of the type 'be hungry1, 'get tired of
84 86 86 89 92 93 1O3 1O4 106 109 111 111 112 113 116 117 117 118
119 121 121 122 126 127 129 13O 130
viii 3.2.8 Some thoughts on verbal prefixes like German be-, OP-, ver-, zer-, and the holistic/partitive/ resultative distinction
134
1
3.2.9 A new look at 'function verb constructs (Funktionsverbgefuge) of the type come to a conclusion—relevance of logico-semantic valence 4.
OUTLOOK 4.1 4.2
5.
140 147
The psychological validity of the concept of a 'conceptual composition plan" for sentences and discourse
. . .
147
The "cyclic approach" for the introduction of related verbs (predicates) based on logico-semantic valence . . . .
148
SUNMARY
153
Zusammenfassung
155
APPENDIX A:
EXPERIMENTAL LANGUAGE TEST IN GERMAN
157
APPENDIX B:
ALPHABETICAL LIST OF VERBS (PREDICATES) ANALYZED AND DISCUSSED IN CHAPTER 3
165
BIBLIOGRAPHY
172
ABBREVIATIONS AND SYMBOLS
179
ix
This work is a modified and bibliographically updated version of my doctoral dissertation, submitted to the Graduate School of Georgetown University in 1975.
It should be pointed out that the research on which
this publication is based has utilized insights from various branches within the field of linguistics, as well as from language pedagogy, with special emphasis on a foreign language teacher's insights into the "joys and agonies' of the foreign language learning process. A debt of gratitude is owed to the many with whom I discussed various aspects of this research, and to my friends who took the time to read, to criticize, and to encourage—especially K. R. Jankowsky, R. Lado, J. Longmire, U. Oomen-Mantell, and H. Vater. A very special thanks to Nora, my wife, for her continuing encouragement and patience. S.R.F.
1.
SOME REFLECTIONS ON THE RELATIONSHIP BETWEEN LINGUISTIC THEORIES AND PEDAGOGICAL GRAMMARS
1.1
On the concept of 'pedagogical grammar'
The rapid and sometimes turbulent developments in linguistic theories during the past decade has initiated a rethinking on the relationship between linguistics and foreign language teaching. Much has been written and said from both the linguistic and the pedagogical point of view, and sometimes it has appeared as if the linguist and the foreign language teacher were each trying to prevent the other from entering his own 'territory' rather than to attain fruitful cooperation.
This confrontation is due, in part, to a mis-
understanding of the respective tasks and goals of linguistics and foreign language pedagogy.
The areas of interest in linguistics are manifold and it
is neither useful nor appropriate to investigate all of them in our context. However, we shall in the theoretical part of this thesis, deal with those areas that seem especially relevant to our investigation. Linguistic theories, in general, are developed without consideration of didactic implications.
A pedagogical grammar, on the other hand, as we see
it, is a (foreign) language acquisition-oriented grammar.
It is not iden-
tical with a general scientific grammar or a simplified version thereof. is not a methodology. textbook.
It
Neither is it a grammar used in a foreign language
Instead, it is a theoretical statement that accounts for the
foreign language acquisition as it occurs in a structured learning situation such as the foreign language classroom.
This notion makes it immediately
clear, then, that input for a pedagogical grammar must come not only from linguistic sources but also from psychology, education, and other fields. Whatever contribution to a pedagogical grammar this thesis will make, such contributions cannot constitute all that is necessary for a pedagogical
graimiar but merely part of the linguistic input into such a grammar. The inmense task of putting together a pedagogical grammar can only be achieved through the interdisciplinary cooperation of linguists from different branches of the field, expert language teachers, textbook specialists (scarce as these may be), information theorists, curriculum specialists, and others. The complexity of such a pedagogical grammar, however, is the main stumbling block for establishing any coherent model. This was also expressed by Jakobovits (1970) and likewise in Corder's major address at the 3rd International Congress of Applied Linguistics in Copenhagen (1972). A pedagogical grammar is language use-oriented, which includes an awareness of the characteristics of verbal strategies in various speech acts, as well as an awareness of the speaker's capability for discrimination and selection of specific syntactic and semantic (lexical) means. Lado's Language Teaching: A Scientific Approach (1964), and his Language Testing (1961) marked the beginning of the objectivization of foreign language instruction. Lado based his observations on verifiable linguistic facts—in pronunciation, syntactic patterns and vocabulary. The performance analysis of a test such as that incorporated in the research here reported can be regarded as part of a language teacher's input into a pedagogical grammar. At the same time, it can be taken as proof of the insight a linguist can gain from observed errors in foreign language learning. For example, the question may be asked: Are certain mistakes generalizable? And why? 1.2
Which linguistic theory?
As will be demonstrated in the theoretical part of this thesis, an eclectic approach to different theories is advisable. Already in 1966, Chomsky made a similar suggestion. Therefore, a considerable part of the theoretical
The hasty transformation of linguistic theories into didactic models is doomed to failure; it has also contributed significantly to a very critical attitude among teachers and students of foreign languages towards linguistic matters. Heringer and Öhlschläger (1973) stress the fact that the incorporation of linguistic theories into a didactic model is not to be considered only a matter of formal adjustment but also scientific work which, in turn, effects linguistic theories.
framework presented for this thesis will be devoted to the investigation of approaches that are centered around sentence patterns (Satzbaupläne). It seems fair to say that this approach is reflected in foreign language textbooks to date.
Pattern practice was derived fron considering major sentence
types as a basis for exercises.
Without entering the discussion on the
degree of usefulness of such exercises, it can be observed that the communicative speech act involves more than just the knowledge of such syntactic patterns.
Every linguistic utterance is an instance of a 'meaningful con-
ceptual plan1 reflecting "shared knowledge1 of the participants in the communicative act.
By producing an utterance, the speaker verbalizes a
concept with the intention of achieving a certain reaction on the part of the receiver; optimally this reaction will result in understanding.
It would be
futile to speculate on the organization of concepts in the mind—we simply do not know enough about that at this time—but it is considered of utmost importance that we accept certain ideas as put forward in semantically-based linguistic theories for a pedagogical grairnar. This aspect was also stressed by Hüllen (1972).
Our remarks on the
relationship of syntax and semantics (2.2.1) will indicate its multidimensionality. dangers.
It is here that 'pattern practice" has its limitations and A semantic sentence structure can often be actualized in more than
one way, and, conversely, one specific surface sentence type can be a manifestation of different semantic sentence structures. at stake.
This is the main issue
It is therefore suggested that sentence-semantic concepts and
case grammar provide an alternate approach as input for a pedagogical gramnar. Recent attempts to incorporate transformational grammar into foreign language (and native language) instruction have not been too successful because they remain primarily within the syntactic sphere. In our context, every utterance is seen as an instance of a "conceptual sentence plan"
(Brekle 1970; Fillmore 1971)·
It is the semantic valence
of the verb that determines the relational constants (semantic cases). It is still possible to work with sentence types—and it would be wrong to abandon valuable work done in this respect—but our sentence types are primarily based on semantic sentence structures, and it is suggested that such an interpretation comes closer to the conceptual level and is closer to communicative competence. It may very well be that eventually new insights
into the relationship between cognition and language structure can be gained from error analysis that is based on our understanding of semantic sentence structures. 1.3
The universality of a pedagogical grammar
It can be assumed from the knowledge we have of the world at large that many semantic relations are intuitively known, as are many compatibility relations between a 'verbal concept1 and its 'arguments'. One of the major difficulties in foreign language learning is the way in which certain 'content values' are expressed in the actual utterance.
To allow compari-
sons between languages it is necessary to use a common terminology for underlying relational concepts. After investigating this matter in some detail in the theoretical part, we will arrive at a relatively small number of relational concepts, e.g. Agent, Object (Identification), Instrument (Tool) etc. These concepts are considered universal in the sense that they can be used for comparisons and contrasts between various languages. No claim is made as to the completeness of such a list.
For the time being it
serves as a useful tool for explaining semantic-syntactic relations and constraints across languages. Innumerable valuable observations on the part of the language teacher are lost due to the lack of a model-like framework.
There is a big dif-
ference between the intuitively felt violation of some semantic rule and the systematic explanation of such a violation with a terminology that permits cross-language reference.
It is rather distressing to observe with
what diverse ideas about the structure of their native language students begin the study of a foreign language. This is to a considerable degree due to different terminology, which too often is limited to morpho-syntactic observations only.
Is it necessary that the student adjusts himself to a
new basic terminology every time he learns a different language, or for that matter, every time he changes English teachers? It is hoped that our critical evaluation will contribute to a better understanding of what ways are available in one particular language, or in several languages, to express certain actions, processes, or abstract images.
In summary, it is observed that the relationship (or better, the interrelation) between linguistics and foreign language pedagogy is an important one.
A pedagogical grammar is not the application of a linguistic theory,
but rather it is a model suitable as a point of departure or source of background information for various tasks in the field of language pedagogy, applicable to foreign languages as well as the student's native language.
2.
THEORETICAL BACKGROUND
2.1
Linguistic valence—a first approximation
Valence is a term that has been borrowed fron the natural sciences. In chemistry, for example, valence is the capacity of an element or radical to combine with another, as measured by the number of hydrogen atoms which one radical or one atom of the element will combine with or replace. (Webster 1971:2017)
Valence is not just a sum (aggregate) of connected elements, but a relationship of dependence or close affinity. The following chapter will give an overview of the use of the concept of valence in linguistic research. 2.1.1
Valence—'a metaphorical terminology'
In a chapter on 'Surface and Deep Grammar1, Hockett introduces the term 'valence1 when discussing the Chinese verb kai, which has a wide range of meaning. The morpheme kai, so to speak, has a 'positive' valence of a special directive kind: it seeks something in the context to seize on as its subject. (Hockett 1958:248)
If that 'something' appears on the surface level at the place required by the rules of grammar, then this kind of valence represents 'surface grammar 1 . If, however, the position of the object remains empty, then the 'unsaturated' valence has to reach further into the context. In this instance valence refers to a level of 'deep grammar'. In a sentence such as I'll drive, the grammatical object is not expressed at all, '... so the valence of the morpheme reaches into the non-speech environment1 (Hockett 1958:249). To symbolize this extension into the extralinguistic level (non-speech situation), Hockett uses the balloon, a graphic device used in comic strips, in combination with IC-diagramming. A unique representation, indeed. But he does not provide us with any definition of valence and evidently uses the term for different levels of analysis.
Although no attempt is node by Hockett to utilize the valence concept for any verb typology, he points out that sing in sing a song and run in run a GOP are instances of different selective categories regarding types of valence, and that such differences are to be found in English at a deep level (1958:250-51).
His concept of 'appositive valence1 in discussing
anaphoric substitutes undoubtedly refers to semantic and possibly pragmatic relationships (1958:254). Hockett's treatment of the valence concept is an indication of the difficulties which one encounters in defining 'linguistic valence'. 2.1.2
Tesniere's concept of valence
It was Lucien Tesniere who introduced the concept of valence to modern linguistics in a comprehensive model for verb typology. In his structural sentence analysis, the verb was the starting point; 'actants1 and 'circonstants1 were considered 'subordines 103).
immediats1 of the verb (Tesniere 1959:
Les actants sont toujours des substantifs ou des equivalents de substantifs ... Les circonstants sont toujours des adverbes (de temps, de lieu, de maniore, e t c . , . . ) ou des equivalents d'adverbes. (Tesniere 1 9 5 3 : l o 2 f f )
In a sense, the valence concept appears in older grammars, e.g., J. W. Meiner, Versuch einer an der menschlichen Sprache abgebildeten Vernunftslehre; oder, Philosophische und allgemeine Sprachlehre, Leipzig, 1781. Meiner uses the concept of 'arguments' and Operators' in a logical sense when he speaks of 'absolute', 'relative' and 'dependent 1 (dreyseitig) predicates (132,143-45) and does not attribute a special status to the subject. J. C. A. Heyse, Deutsche Grammatik, Hannover, 19O8, and 0. Behaghel, Deutsche Syntax, Vol. 2, Heidelberg, 1924, distinguish between 'absolute' verbs, requiring only a subject, and 'relative' verbs, which require at least one complement. K. Bühler, Spzachtheorie, Jena, 1934, recognized ... daß die Wörter einer bestimmten Wortklasse eine oder mehrere Leerstellen um sich eröffnen, die durch Wörter bestimmter anderer Wortklassen ausgefüllt werden müssen. (1934:173) Bühler uses 'Leerstelle' in a connotational sense.
The characteristic of a verb to require a certain number of actants is compared by Tesniere to that of the structure of the atom and is called valence. Based on the number of actants, Tesniere distinguishes the following major verb-classes: 1)
Verbs without an actant ('avalent1 verbs) e.g., il pleut, es regnet 'it's raining". By not recognizing il as actant, Tesniere applies valence criteria at different levels but does not supply us with reasons for distinguishing between 'avalent1 verbs (content plane) and 'monovalent' verbs (expression plane), except: Quant un sujet apparent qui se trouve dans certaines langues ( f r . il pleut, il neige, il vent; all. es regnet) ce n'est en realite, comme nous 1'avons vu que le marquant de l'anontif singulier du verbe. (Tesniere 1953:239)
2)
Verbs with one actant ('monovalent' verbs) Tesniere distinguishes between different types of actants, such as: 'prime actant 1 ... celui qui fait 1'action; "second actant' ... celui qui support l·1action; 'tiers actant 1 ... celui au benefice ou au detriment duquel se fait 1'action. (1953:lO8ff)
3)
Verbs with two actants ('divalent' verbs).
4)
Verbs with three actants ('trivalent' verbs).
Candidates for valence consideration are only the subject, the direct object and the indirect object; prepositional objects and adverbials ('circonstants1) are not considered, e.g. ... wohnen 'habiter', lequel se construit avec une preposition de lieu, que introduit egalement un circonstant. ... (Tesniere 1953:269)
The subject in a sentence has no special status for Tesniere.
Looking back on Tesniere's work we can evaluate it on the basis of how it stijtulated further research in valence theory, especially through its shortcomings. 1)
This influence may be summarized as follows:
The level at which Tesniere carries out his verb typology is unclear, to say the least.
In discussing ava-
lent verbs he considers actant on one level, as mentioned above; on the other hand his morphological criteria for valence classification indicate that his actants are formal representations on another, the expression level. The decision as to whether valence is a property of the expression plane and thus observable in distributional syntactic data of a particular language, or whether it is a characteristic of the semantic level, or, even more abstractly, of some "conceptual sentence level1, is of crucial importance. Tesniere seems to extend the valence concept even to this abstract level when he speaks of la nouvelle valence ä marquant zero ... une valeur causative trivalente ... Caesar pontem fecit. (Tesniere 1953:272)
where the causative never appears on the expression level. 2)
Tesniere's distinction between 'actant1 and 'circonstant', based on morpho-syntactic criteria, appears to be too narrow, since no provision is made to allow prepositional phrases as instances of actants. Thus, for example, the prepositional phrase in Neu York in
Tesniere separates on morphological grounds (with approximately the same meaning) such sentence pairs as: Alfred wohnt in diesem Haus. and Alfred bewohnt dieses Haus. "Alfred lives in this house.' not accepting the prepositional phrase in diesem Haus as a necessary 'actant 1 of the verb wohnen. (Tesniere 1953:269)
10
He lives in New York would be considered a circonstant, even though it is required by the verb, if it is to be understood in the sense of reside and not to be alive (see also footnote 2, p. 9 ) . 2.1.3
The valence concept (Valenzbegriff) in German linguistics
2.1.3.1
Different interpretations of valence
Until recently, when linguists became interested in textual analysis, one of the main concerns of linguistic analyses was the sentence patterns (Satzmodelle) .
In the American structuralist tradition, the subject-predicate
dichotomy as well as substitution patterns played a significant part in establishing sentence (clause) types (see Fries, Hockett); even Harris1 kernel sentences (1957) are not intended to reflect semantic manifestations. At first sight, the basic patterns (Grundmodelle) in Johannes Erben"s Deutsche Grammatik— Ein Abriß (1972) appear to be based on the same structuralist considerations as those of the American linguists. considers the verb as the nucleus (Aussagekern).
Erben
So his basic patterns
rely on the selection of the verb and its valence (Wertigkeit): Von seiner syntaktischen Wertigkeit ("Fügungspotenz")— man kann sie geradezu mit der Valenz des Atoms vergleichen—hängt es wesentlich ab, welche und wie viele Ergänzungsbestimmungen im Vor- und Nachfeld des Verbs auftreten und das Satzschema ausgestalten. (Erben 1972:246)
According to the number of objects (Ergänzungsbestimmungen) Erben distinguishes sentences with one, two, three, or four objects; these basic patterns are then subdivided into realizations (Spielarten) on the basis of the syntactic form (Strukturbilder) of the objects. In his objects Erben does include predicatives, obligatory prepositional phrases, and obligatory adverbial phrases. He thus extends the concept of valence on the syntactic level beyond that of Tesniere. Although the distinction Obligatory/facultative1 is made (Erben 1972:248), no criteria are provided for making such a distinction.
11
In his 1972 edition, Erben is more careful in attributing a certain content (Inhaltswert) to basic syntactic models (1972:258).
He stresses
the need to consider sentence-semantic aspects in a detailed description of sentence models.
This fact is illustrated when discussing the various
Strukturbilder and their Spielarten, where such semantic relations as Agens, Patiens, and Zielgröße are used and compared to Fillmore's deep case relations (1972:144).
Ihe detailed semantic discussion of various verb groups
and their Ergahzungsbestimmungen offers a wealth of insight into the intricate semantic-syntactic interplay but it does not attempt any systematic treatment of such (inter)relations. Hennig Brinkmann in Die deutsche Sprache— Gestalt und Leistung (1971) discusses the concept of valence in some detail. Tesniere in his interpretation of valence.
He appears to follow
It is true that Brinkmann, like
Tesniere, does not consider prepositional objects and adverbials potential actants (Mitspieler).
However, Brinkmann's special treatment of the subject-
predicate relation, and consequently the exclusion of the nominative case as a valence position, constitutes a fundamental difference between his and Tesniere's concept of valence (see also Fourquet, 1976).
This fact must be
stressed even though Brinkmann in passing mentions this difference in the 1971 edition in a footnote: Anders als es hier geschieht, gibt Tesniere dem Subjekt keine Sonderstellung. (Brinkmann 1971:21O, footnote 1)
Considering three positions—accusative case, dative case and the singledout nominative case—Brinkmann arrives at a 'syntactic verb-sphere typology'.
By focusing on the verb's sentence-value (Satzwert), he distin-
guishes eight verb classes ranging from zero-place verbs and limited oneplace verbs to three-place verbs (Brinkmann 1971:212-19).
The criteria
Brinkmann uses to establish these verb classes are drawn fron syntax, semantics, pragmatics and philosophy. Brinkmann distinguishes between 'certain general relations' (bestimmte allgemeine Beziehungen), which he attributes to the nominative (Kasus der Identität), accusative (Kasus der Alterität), and dative (Kasus der Finalität); and 'idiosynchratic relations' (spezielle Beziehungen), which hold for prepositional phrases. Genitive appears to be an ambiguous case (1962:514).
12
The difficulty in defining the "content value1 of a sentence is evident in Wladimir Admoni's logico-grammatical sentence types. Der deutsche 'Jpraahbau
He maintains in
(1970:231) that this terminology reflects best
... die Tatsache der Wiederspiegelung von objektiven Sachverhalten—durch das Medium des menschlichen Denkens— in der Struktur des Satzes. (Admoni 1970:231)
and indicates the similarity of such logico-grammatical sentence types in different languages.
The number of his (originally five) main sentence
types and their 'generalized content' (verallgemeinerter Bedeutungsgehalt) has been extended to twelve (Admoni 1970:235-42). The finite verb in German has, according to Admoni, at least one actant; ... eine 'Leerstelle' der obligatorischen Fügungspotenz, namentlich die der Fügung mit dem grammatischen Subjekt, von welchem der verbale Prozeß ausgeht. (Admoni 1970:158)
and thus he indicates the importance he attributes to the surface subject as part of the sentence nucleus (struktureller Satzkern) (Admoni 1972:72). Admoni admits the usefulness of dependency relations (Abhängigkeitsbeziehungen) as one of several devices to describe syntactic relations, but they must not be the basis for an independent syntactic theory. 4
Admoni uses the term 'Fügungspotenz' (valence) in a very general sense, not restricted to the verb: Jeder Redeteil enthält also in sich eine ganze Reihe von Fügungspotenzen, die bei seiner Einschaltung in den Satz als Ausdruck der vom Redenden beabsichtigten Bedeutungsfüllung aes Satzes und unter dem Einfluß von Kontext und Situation zum Teil aktualisiert werden. (Admoni 197O:8O)
5
Admoni 1974:72: 'Als ein Z u g r i f f , der in Verbindung mit anderen Zugriffen das Wesen der syntaktischen Beziehungen im Satz zu klären h i l f t , ist die konsequent und tief durchgeführte Bestimmung der Abhängigkeitsbeziehungen zwischen den Wörtern im Satz nicht nur zulässig, sondern sogar unentbehrlich. Aber wenn dieser Zugriff verabsolutiert wird und als eine besondere Grammatiktheorie zur Erschließung der Satzstruktur zu dienen hat, so führt er unverweigerlich zur Verkennung der wahren Struktur des Satzes. Zu solcher Verkennung gehört vor allem die völlige Degradierung des Subjektnominativs und des Prädikativs zu Komponenten, die durch einseitige Abhängigkeit vom finiten Verb gekennzeichnet sind und nicht dem strukturellen Zentrum des Satzes zugerechnet werden.' Admoni claims to have incorporated the concept of valence in 1935 although without using the terminology (1974:34).
13
In their treatments of valence with regard to sentence models, Tesniere, Erben, Brinkmann, and Admoni attempt to obtain a limited number of basic sentences that constitute what one might call the 'grammatical core' of man's linguistic ability.
These sentence models, then, represent the
foundation upon which infinite varieties of actual, more complex sentences, can be built for communicative purposes.
The methodological approaches in
describing such sentence models differ considerably, although they all take the syntactic form of the sentence as a starting point. 2.1.3.2
Heibig's concept of valence
The first publication of Wörterbuch zur Valenz und Distribution deutscher Verben by Gerhard Heibig and Wolfgang Schenkel (1969) brought valence to the attention of the applied linguists, especially those involved in teaching German as a foreign language.
Helbig follows Tesniere in that he con-
siders the verb as the structural center of the sentence and thus abandons the traditional subject-predicate distinction.
Syntactic valence is de-
fined as ... die Fähigkeit des Verbs, bestimmte Leerstellen um sich herum zu eröffnen, die durch obligatorische oder fakultative Mitspieler zu besetzen sind. (Helbig/Schenkel 1973:49)
These vacant positions (Leerstellen) can be filled by objects (including certain prepositional phrases), subject, predicatives, and some adverbial phrases. To arrive at the syntactic minimum, Helbig and Schenkel (hereafter referred to as HuS) apply the deletion test, the syntactic valence of the verb.
which allows them to establish
While the necessary members, delimited
by the syntactic valence, are tied to the verb and limited in number and kind, the "free adjuncts' (freie Angaben) have almost no restrictions as to number and the sentence type to which they can be added (HuS 1973:33-34). The 'deletion test 1 and the 'permutation test 1 were adopted by Helbig and Schenkel (herein referred to as HuS) from H. Glinz, who independently arrived at criteria very similar to those of Fries/Lado for the structural description of sentences. Incidentally, Glinz also considers the verb the center or axis of the sentence and calls the verb 'Leitglied 1 (Glinz 1962:96).
14
HuS make a further distinction, dividing the necessary actants into obligatory and facultative actants (obligatorische und fakultative Valenz), a concept that will be discussed in more detail later (2.2.3).
Some of HuS's
examples may illustrate the tripartite distinction they make: facultative, free.
obligatory,
(1)
Mein Freund wohnt in Dresden.
obligatory
'My friend is living in Dresden.' (2)
Er wartet auf seinen Freund.
facultative
'He is waiting for his f r i e n d . ' (3)
Er aß sein Brot in der Schule.
free
'He ate his sandwich in school.' (HuS 1973:34)
Each of the prepositional phrases in ( 1 ) , (2), and (3) has a different syntactic status, according to HuS. Only in der Sohule, in sentence (3), stands in no valence relation to the verb. The deletion test separates obligatory actants, as in ( 1 ) , from facultative (2) and free (3). The obligatory actants together with the verb constitute the minimal grammatical sentence. The prepositional phrases in (1) and (2) are considered 'close verbal complements ' (EV, enge Verbergänzungen), which are relevant for verb classification. The prepositional phrase in (3) is a 'free verbal complement1 (FV freie Verbergänzung). While EVs are constituents of the verb phrase, FVs are immediate constituents of the sentence. To distinguish facultative actants from free adjuncts, a paraphrase test employing an adverbial clause is used. This is considered a 'deep structure' test by HuS. It is possible to change (3) to (3a): (3)
Er aß sein Brot in der Schule.
(3a)
Er aß sein Brot, als er in der Schule war. "He ate his sandwich while he was in school."
thus establishing in der Schule as a IV. It is not possible to do the sane with (1) (1)
Mein Freund wohnt in Dresden.
(la)
*Mein Freund wohnt, als er in Dresden war.
15
Obligatory actant positions must be realized in the actual sentence while facultative actants may be vacant in the actual sentence, depending on such factors as situational context and speech context (HuS 1973:53). All FVs are seen as reduced sentences and reflect a predicate-logic relation different from that of EVs.
Possible restrictions in the distribution
of FVs are attributed to inherent semantic markers of the verb and their incompatibility with certain FVs; but these are not dealt with within syntactic valence. HuS refer to different levels of valence (syntactic, semantic, and logical), but claim to have based their dictionary on syntactic valence only.
This, of course, is a reflection of semantic facts: Es besteht kein Zweifel daran, daß eine solche direkte Beziehung in vielen Fällen besteht, daß die Valenz-und Distributionsbeziehungen im allgemeinen ein formaler Reflex semantischer Gegebenheiten sind. (HuS 1973:60)
In comparing prepositional objects (präpositionales Objekt) with adverbials with prepositions (präpositionales Adverbial), HuS observe that prepositions in the prepositional object are syntactically verb-governed, usually restricted to one particular preposition (as in (4) below), and are semantically 'empty1; while the prepositions in adverbial phrases are not verb-governed (as in (5) below), and carry part of the semantic information of the entire phrase, and are frequently substitutable: (4)
Er wartet auf den Freund. "He is waiting for his f r i e n d . '
(5)
Er wartet
\
.
/ dem Bahnhof.
'He i s waiting < , , . - ,
) t h e station. '
16
Verb-governed prepositional objects are always obligatory actants ( 4 ) ; but necessary actants are not always verb-governed ( 1 ) : Regierte Glieder sind immer valenzgebunden; aber—wie die notwendigen Adverbialbestimmungen zeigenvalenzgebundene Glieder sind nicht immer regiert. (HuS 1973:44)
By using only prepositional phrases, and taking substitutions of the preposition as criterion for a FV/EV distinction, HuS do not cover FVs and EVs in the form of adverbs.
According to HuS, adverbial phrases as nec-
essary actants can be of two kinds: 1)
The occurrence and kind of adverbial is determined by the valence of the verb, e.g.:
2)
wohnen
(place; locality)
"to live, reside 1
legen
(direction)
"to put'
aussehen
(modality)
"to look like 1
The occurrence, but not the kind of adverbial, is determined by the valence of the verb, e.g.: sich ereignen
(location, time, cause)
'to take place 1
There is a choice between several adverbials, but at least one is necessary. Two further tests, permutation and negative transformation, are employed to distinguish EV and FV.
In German, the closer the affinity of an actant
to the verb, the closer the actant appears to the end of the sentence.
This
means that an 'adverbial phrase of first degree" (EV) would follow one of 'second degree1 (FV), in a normal, non-contrastive sentence.
The following
example is given: (6)
Du hast das Buch am Vormittag dorthin gelegt. 'You put the book there in the morning. 1
Further criticism on this matter is voiced by L. Götze in a review of HuS's Wörterbuch in: Zielsprache Deutsch 1973/3,137-38.
17 (6a) (6b) (6c)
?Du hast das Buch dorthin am Vormittag gelegt. Du hast das Buch dorthin gelegt. *Du hast das Buch am Vormittag gelegt.
(HuS 1973:47} While dorthin is the obligatory part of the valence, con Vormittag is a FV Q
and can therefore be eliminated. The second additional test is the 'negative transformation test': (7)
Ich lege das Buch nicht dorthin. 'I don't put the book t h e r e . '
(7a) (8)
?Ich lege das Buch dorthin nicht. Ich treffe
ihn dort nicht.
d o n ' t meet him there. 1 (8a)
Ich treffe
ihn nicht dort.
The fact that sentence negation precedes the obligatory adverbial phrase is the cause for the awkwardness of sentence 7a; dorthin is thus, on syntactic grounds, established as part of the verb valence. Let us now consider the application of HuS's understanding of valence in their valence dictionary: Verbs are interpreted on three levels.
9
Level 1—Nunber of obligatory and facultative actants (in brackets); added up they constitute the quantitative valence of the verb, e.g.: l warten
/ » ) _ - > (harren, erwarten) 'to wait' (HuS 1973:432)
HuS follow Bierwisch (1965) with this and the negative transformation test. Bierwisch, when discussing constituent structure, speaks of degree of affinity to the verb; he uses besides 'nicht 1 emphatic words (Affirmationspartikel) such as ' d o c h ' , 'bestimmt', and says Wenn sie zum ganzen Satz gehören und nicht zu einem einzelnen Glied, dann haben sie einen festen Platz vor dem Glied, das am engsten zum Verb gehört, ... (1965:36) It is evident, and Heibig himself points it out ( 1 9 7 1 a ) , that this three-level division is a combination of dependency, following Tesniere, the structural grammar of Z . S. Harris, and Chomsky's generative grammar. Borrowing concepts from different linguistic models and applying them at different levels causes considerable confusion in HuS's valence model. For additional comments on this problem, see also Ballweg, et al. (1971: 102 f f ) .
18 Level 2—Qualitative description of the actants in terms of the syntactic environment of the verb, e.g.: 1
2 warten where NS
Sn = noun (nominative) pS = preposition + noun j n L. = clause introduced by daß,ob,w ·* that, if, WH Inf = infinitive with 'to' / = alternate possibility
Level 3—Qualitative description in terms of semantic specification of possible actants, e.g.:
1 2 3 Sn
*-1. 2. 3. 4.
+Anim (Der Freund, der1 Hund wartet.) Abstr (as Hum) (Das Ministerium wartet.) -Anim (Fahrzeug) (Das Auto wartet.) Abstr (Die Arbeit wartet auf uns.)
p = auf
pSa
» keine Selektionsbeschränkungen (prepositional phrase in the accusative case; no selectional restrictions)
NS '·
» Act (Er wartet darauf , daß ich komme.) (action)
Inf
»-Act (Er wartet darauf, gefragt zu werden);
A footnote to the entry warten states that: when with V-j Sn »Abstract then pS is obligatory and NS is not possible (HuS 1973:433). Although level three states the semantic environment, it is, according to HuS, not to be interpreted as semantic valence: Die semantische Valenz ergibt sich aus den Verträglichkeitsbeziehungen zwischen den Bedeutungskomponenten (Semen, Noemen, Bedeutungsmerkmalen u . a . ) von Verb und Aktant; die Selektionsbeschränkungen des Verbs sind nicht mit ihr identisch, sondern eher ein syntaktischer Reflex dieser semantischen Valenz. (HuS 1973:53)
19
HuS observe that if variants of a verb are manifested already on level 1, the degree of difference in meaning is greater than if they appear on level 2 or 3.
Thus, for example, the verb V~ warten has the valence index
2
and the meaning 'to take care of (care for, service)'. Deletion of facultative actants does not change the 'denotative meaning1 of the verb, only the semantic aspect (influence of context).
If an ob-
ligatory actant is vacant, we have, according to HuS, a case of ellipsis (no influence from context, since only one filler of the vacant position is possible), but no difference in 'denotative meaning1. In the 1973 edition of HuS's valence dictionary, about 5OO German verbs are analyzed.
The selection was made in accordance with the frequency of
use in the spoken language and the difficulty in learning, based on observations with students of German as a foreign language.
Many verbs of
cortrnon usage and with seemingly simple lexical structure that are frequently introduced in the first weeks of instruction, are, in fact, given a very complex description in HuS's dictionary.
While these verbs often have
several variants, many verbs with low frequency have no variants and a short description.
The observation of 'semantic load", then, is of crucial import-
ance in the discussion of degree of difficulty in learning verbs.
If this is
the case, a most carefully prograimed cyclic approach to vocabulary selection, especially verb treatment, must be undertaken in the preparation of foreign language textbooks. Any systematic behavior of verb subclasses based on valence or case grammar may be of utmost significance (see 4 . 2 ) . The relative complexity of certain verbs is reflected in the sentence models.
Based on HuS's valence concept, we find ten major sentence models
in Deutsche Grammatik:
Ein Handbuch für den Ausländerunterrioht by Gerhard
Heibig and Joachim Buscha (1972).
These sentence models, containing verbs
In an article "Zu einigen Spezialproblemen der Valenztheorie" (1971b), Heibig discusses the difference between facultative valence and ellipsis: Das hauptsächliche Kennzeichen der Ellipse besteht darin, daß eine völlige Bedeutungsgleichheit zwischen der vollständigen und der reduzierten Form besteht. (Heibig 1971b:276) Erben (1972:248) speaks of a 'superficial polyvalence 1 (scheinbare Polyvalenz) when discussing obligatory and facultative actants. Contrary to HuS above, Erben considers ellipsis as an instance of an eliminated facultative actant.
20
without an actant and verbs with up to three obligatory actants, are subdivided into no less than 97 subtypes.
They illustrate the wide variety of
possibilities in basic German sentence structure.
It is very clear, how-
ever, that these subtypes are not based on semantic criteria (Helbig/Buscha 1972:551, 554-58). 2.1.3.3
Engel's 'sentence models' (Satzbaupläne)—Valence research at the Institut für Deutsche Sprache^
Ulrich Engel and his associates have been working with the valence concept and have arrived at 'sentence composition plans' (Satzbaupläne) very similar to those of Helbig/Buscha.
The main objective of the research of Engel
et al has been the application of sentence models in the teaching of German as a foreign language and in the description of the structure of German to native speakers of the language.
The following remarks by Engel must be
seen with this applicational goal in mind: Die Klasse 'Verb' zerfällt in disjunkte Valenzklassen, die durch Zahl und Art ihnen spezifischer Ergänzungen definiert werden. (Engel 1970:371)
Since Engel is concerned with basic sentence models, the verb (Regens) and its dependents (Dependens) are of primary interest.
Helbig's criteria--
deletion, substitution, and paraphrasing—are, for Engel, not sufficient to establish basic sentence models.
He uses anaphora,
and, depending on
the kind of anaphora (personal pronoun, adverbs, da + preposition), he distinguishes: Casualia, Praepositionalia, Adverbialia, Neutralia, Verbalia, and the remaining group (Engel 1970:366ff).
Ten 'complements' (Ergänzungen),
with the following codes, are distinguished: G
Nominativobjekt
(nominative object)
i
Akkusativobjekt
(accusative object)
's.
Genitivobjekt
(genitive object)
11
The research in Mannheim carried out under the project title "Gruridstrukturen der deutschen Sprache" is still in progress. Detailed progress reports appear in the monograph series Forschungsberichte des Instituts für Deutsche Sprache. Basic concepts about the valence dictionary are presented by H. Schumacher ( 1 9 7 2 ) .
12
Anaphora is defined as the most abstract member of a certain paradigm, its function being to refer to something already mentioned (antecedent) (Engel 1970:366). Engel cites K. Bühler as the source for his anaphora concept (Bühler 1934:121, 3 8 5 f f ) .
21
3
Dativobjekt
(dative object)
4
Präpositionalergänzung
(prepositional object)
5
statische Adverbialergänzung
(Stative adverbial phrase)
6
Richtungergänzung
(directional object)
Numerabile
(numerals
l \ als Ergänzung
Comparabile I
> (comparatives
•l l
/
Verbalergänzung
as complements)
(verbal complement)
These complements enable us to subclassify verbs into disjunctive groups according to their dependency relation with specific verbs.
With various
subtypes, Engel obtains 3O sentence composition plans, e.g. O136
Monika bringt ihrem Vater einen Brief an den Zug. O 3 1 6 (Engel 1970:376) 'Monika is bringing a letter to her father in the train. 1
The code numbers (e.g. 0-1-3-6 above) include obligatory and facultative valences but exclude free adjuncts, according to Engel.
If the same verb
has two different codes, we have an instance of hononymy.
Engel admits that
this is an indication of a difference in meaning (and meanings, contents, are, for Engel, the main goal of linguistic investigation) and he suggests treating these hononyins as instances of separate verbs. Although the tests mentioned above are sufficient to identify homonyms, Engel does not use them to arrive at semantically different sentence structures. This appears inconsistent with Engel's general interest in the 'content value' of sentences. Zwei homographe Verben sind als verschiedene Lexikoneinheiten zu betrachten, wenn und weil sie verschiedene Valenz und damit verschiedene Satzbauplär.e haben. Der konstatierte Bedeutungsunterschied darf in den gegebenen Zusammenhängen nur als (wenngleich willkommenes) Akzidens gewertet werden. (Engel 197O:38O)
From Engel's ten complements only those with codes 5, 6, 7, and 8 can be connected with certain meaning elements, while the others cannot be semantically defined; their 'content value' changes from verb to verb.
Engel,
22
therefore, suggests attributing semantic characteristics to the verb, and he distinguishes between valence on the morphological and the semantic level: Da das Verb außerdem die Form des Objekts festlegt, besteht Valenz als Eigenschaft des Verbs aus zwei Merkmalen: einem morphologischen (das Kasus u s w . , damit die Klasse der Ergänzung festlegt) und einem semantischen (das der Ergänzung ein Bedeutungselement zuordnet, das aus der Ergänzung selbst nicht erschlossen werden k a n n ) . Zwischen den beiden Merkmalen besteht kein weiterer Zusammenhang, die Zuordnung ist zufällig. (Engel 1970:383)
According to Engel, there is no systematic relationship between these two levels; one cannot correlate a certain "content1 with a certain surface case (Ergänzung). In this context, Fillmore's attempt to base each actant on a 'deep prepositional phrase1 (1966a) is mentioned by Engel as striking and original but not developed enough: Ein sehr origineller Versuch, allerdings auf das Englische beschränkt, wurde von Fillmore unternommen. Dieser amerikanische Linguist schreibt jedem "actant", jeder Art von Ergänzung also, ursprünglich eine bestimmte Präposition mit je eigener Bedeutung zu; Termini wie Agentiv, Benefaktiv usw. sollen diese Bedeutung andeuten ... Das bestechende Verfahren ist allerdings in seiner Tauglichkeit für das Englische noch nicht hinreichend geprüft; Untersuchungen für das Deutsche fehlen fast völlig. (Engel 1970:382)
It will be one of the major tasks of this thesis, however, to investigate the feasibility of combining aspects of valence with insights from case grammar for application in a pedagogical grairmar. Let us briefly consider Engel's position on obligatory and facultative valence.
A facultative actant is defined as one whose actualization (pre-
sence) in the sentence is irrelevant for the gramnaticality of the sentence; on the other hand, an obligatory actant must be present.
The term 'faculta-
1
tive and 'obligatory' must not be confused with "necessary" (notwendig) and 'omissible' (weglaßbar), the latter being used in a comnunicative sense: Von einer bestimmten Mitteilungsabsicht aus sind gewisse Elemente notwendig oder weglaßbar, und dies steht in keinerlei Beziehung der Grammatizität. Ein kommunikativ notwendiges Element ... ist oft grammatisch fakultativ; und selbst grammatisch obligatorische
23
Elemente können bisweilen aus kommunikativen Rücksichten weglaßbar sein ... (Engel 1970:372)
Engel finds it necessary to make a distinction between such dichotomies as 'specific'/'free', 'obligatory'/'facultative', and 'necessary'/'omissible '.
It is suggested that complements with codes 7 and 8 are mostly
instances of obligatory valence; and all other conplements are, depending on trie governing verb, obligatory or facultative.
The valence of a verb is,
therefore, determined by the sum of its obligatory and facultative actants (Engel 197O:379-8O).
Like Helbig, Engel puts facultative actants in
brackets, e.g. warten
O(4) Alle warten auf das Wunder. 'All are waiting for the m i r a c l e . '
From this brief survey on linguistic valence it is evident that the valence concept is applied at different levels: possibly pragmatic.
syntactic, semantic, and
This fact has contributed to considerable confusion
in the interpretation of linguistic valence.
None of the linguistics men-
tioned used valence unambiguously on one particular level.
This strongly
suggests the necessity of considering valence on more than one level. Major criteria to be used in the evaluation of valence-related matters as they are discussed in various linguistic models, are: 1)
At what level(s) of analysis does the particular model function?
2)
What tests are used in a particular model to support its claims regarding valence distinctions?
3)
Are the descriptive devices used in a certain model of implicational value to or applicable in a pedagogical grammar?
Concerning the distinction 'obligatory'/'free'
(which will be further
discussed in 2.2.3 and 2.2.3.1) on syntactic grounds the elimination test, the substitution test and the negative transformation test are considered valid criteria for establishing obligatory actants.
HuS's terminology in
their valence dictionary for various possible morpho-syntactic valence realizations is accepted for use in the applicational section of this thesis (see 2.1.3.2, under qualitative description on level two, for further details on HuS's terminology; p. 18.)
24
Engel's observation that no systematic relationship exists between the syntactic and semantic level cannot be shared by us.
To be sure, there is
no 1:1 relationship, but there are certainly 'more or less favored' syntactic forms for such notions as location, benefactive, and agent. 1
A confu-
1
sion of Obligatory with 'necessary , as indicated earlier, must under any circumstances be avoided if valence is to be defined within a syntactic framework. 2.1.3.4
Vater's revised generative dependency model
Returning to Tesniere's concept of valence, as outlined in 2.1.2 above, we can see that the sentence consists of a network of structural connections which, in turn, establish certain dependencies between parts of the sentence. The question of the priority of semantics over syntax or vice versa shall not enter our discussion here.
It is hypothesized, however, in view of the
applicational side of our research, that sentence-semantic considerations supply a viable representation of psychological-cognitive concepts; and, therefore, carry potentials for a communicative-functional description of language. Based on Tesniere's work, as well as studies by Hays (1964), Baumgärtner (1970) and Robinson (1970), Vater developed a generative dependency grammar. Vater adds to Robinson's dependency model the valence of the verb (Helbig/ Schenkel, 1973); and on the semantic level he adds aspects of Fillmore's 'cases' (Fillnore 19 68, 1971).
Vater states:
... each governing element in a dependency grammar not only has to have the number and categorial status of its dependents (i.e. its valence) specified, but also their semantic properties. (1975a:134)
In other words, semantic as well as syntactic restrictions control the combinability of two elements in a dependency relation.
Dissatisfied with
Fillmore's treatment of frame features, Vater suggests abandoning the
25
concept of 'case1 in favor of a decomposition into features. achieved by using two types of 'contextual features':
This is
'valence features',
specifying the number and categorial type of dependence; and 'complementation features', specifying the semantic function of the governed elements. This semantic function is, as we see it, a relational function between the governed element and the governing verb. An example may illustrate this concept: . , +
write: ite:
., ,, „, +Human ,Ν,Η,Ν, + __ +Agent
+
+Human : — +Expericnccr
+
-Animate | — +Patient
(Vater 1975a:139)
The 'complementation features' rather closely correspond to FiUmore's 1
cases', but they do not include inherent semantic features of Ν, such as
1
human' or 'animate'.
These latter features are entered above the 'comple-
mentation features' as "specifications of the (simultaneous) environment1 (Vater 1975a:138).
Inherent semantic features of verbs such as 'Act(ion)',
'Com(munication)' are entered separately.
In doing so, Vater obtains a
classification of verbs including 'the most relevant properties of verbs' (1975a:142), ,
Γ
_
„. . , .
+Anim
push + Act, - Com, +· N,N(N). + ^ | — ' — +Agent
see
[
- Act, + Com, +
N,N/T, +
+An:Lm
+Exp
+
+
+Concr /
+Concr\|
(+ 1 — +Pat V — +InstrJI
+ Pat J
(Vater 1975a:142)
Two advantages in using contextual features are, according to Vater, that concepts such as Agent and Instrument are not stated as inherent qualities of nouns, and that these contextual features are not semantically defined in terms of noun features such as "animate1 and 'human'. Vater's model, when compared with the previous models using linguistic valence, demonstrates the need to incorporate semantic information in a formalized manner.
In this respect, the lexical entries for verbs as
illustrated above constitute a significant step forward.
The usefulness
of introducing inherent noun features as 'specifications of the (simultaneous) environment" of "complementation features' is open to question. We will make seme alternate suggestions in Section 2.2.4.3 regarding the specification of 'complementation features'.
26
2.2
Thoughts on the isolation of irorpho-syntactic units and their relation to semantic concepts; morpho-syntactic valence versus logico-semantic valence
2.2.1
Semantics and syntax Expression and content are solidary—they necessarily presuppose each other. An expression is expression only by virtue of being an expression of a content, and a content is content only by virtue of being a content of an exprestion. (Hjelmslev 1961:48)
The above statement on the solidarity of expression and content may be surprising. And yet, even if we take 'meaning1 as central in the linguistic theory, semantics (content) is only part of the total system of communication. Considering the transformation of configurations of ideas to sound/ writing, or vice versa, no position is taken here stressing unidirectionality as is emphasized in 'generative semantics1 (semantic base) or in 'interpretive semantics1 (syntactic base). We do, however, consider the content side as the starting and end point in our investigation, realizing that 'well-formedness conditions' in the expression side are significantly different from those in the semantic (conceptual) domain. Detailed structural descriptions of languages give evidence of language-specific raorphosyntactic rules that have to be applied to account for syntactic wellformedness . Our major concern is the function of sentence constituents such as noun phrases, predicates, and adverbials, and how such functions as subject, object, and complement relate to semantic concepts. The morpho-syntactic well-formedness depends largely on the combination and ordering of constituents, which in the case of English is strictly limited. Violations of these rules result in a lesser degree of grairmaticality. If one considers the predicate to be the nucleus of the sentence, it becomes immediately apparent how closely associated sentence patterns are with a particular class of verbs. This is a well-known observation which was, and still is, the main issue in developing foreign language materials. Semantic well-formedness does not necessarily guarantee syntactic well-formadness. It can be assumed that the concept of 'someone being very tired1 could be rendered in English as *'John very tired'; the syntactic requirements of English, however, make the presence of a predicate obligatory, even if it does not express any "content1.
27
An instance of "semantic emptiness' can be observed in meteorological utterances such as It is raining, where the subject position is filled by it,
due to the requirements of English syntax that a sentence should have
a surface subject.
What we are concerned with in describing semantic well-
formedness conditions may be called 'predication analysis' (Leech 1974: 134).
The example above, It is raining, may also support the assumption
that the predicate is the central part of every 'predication' since it is an instance of a "no-place predication', i.e. having no argument, just like It is hot. These examples indicate that there is no one-to-one relation between semantic and syntactic units in both directions, so we ... must be wary of expecting these logico-semantic units to have any straightforward correspondence with syntactic units. (Leech 1974:129)
It is this 'many-to-one1 mapping relationship between semantics and syntax that brings about, according to our view, a multitude of difficulties and potentials for errors in foreign language learning. In order to investigate the acceptability of an utterance we also have to consider 'expression rules' (Leech 1974:181) that link semantically well-formed prepositional concepts to syntactically well-formed sentences. 2.2.2
Isolating syntactic constituents
Before we examine correlations between semantic elements, such as arguments and predicates, and their syntactic counterparts, it is necessary to reflect on criteria used for the delimitation of syntactic constituents (syntagmas). The distinction between syntagmatic and paradigmatic relations has been extensively used in structural linguistics and defined thus: By virtue of its potentiality of occurrence in a certain context a linguistic unit enters into relations of two different kinds. It enters into pa rad igma tic relations with all the units which can also occur in the same context (whether they contrast or are in free variation with the unit in question); and it enters into syntagmatic relations with the other units of the same level with which it occurs and which constitute its context. (Lyons 1968:73) 1
Leech makes the quite useful distinction between 'predication 1 as the semantic correlate of a sentence and 'predicate' as the semantic correlate of the verb (1974:135). We follow this concept in our research.
28
The substitution test comprises the primary criterion, supplemented by the permutation test.
Brinker (1972) distinguishes four theoretical possi-
bilities for the status of a syntactic constituent: 1)
permutable and substitutable
2)
not permutable and not substitutable
3)
permutable, but not substitutable
4)
not permutable, but substitutable
Brinker uses these criteria for German only; we would like to use them in a wider framework and will therefore test their applicability to English. The operations above must not result in any change of "information content' in the sentence. 1)
Permutable and substitutable: ('·)
The lady from New York arrived with the eight o'clock train.
This sentence could undergo the following permutation: (9a)
With the eight o'clock train the lady from New York arrived.
(9b)
The lady arrived with the eight o'clock train from New York
(9c)
From New York the lady arrived with the eight o'clock train.
but not:
since (9b) and (9c) would violate the dependency relation in which from New York modifies the lady.
The syntactic constituents of (9) at the clause
level would then be the lady from New York/'arrived/with the eight 'cloak train; the lady from New York could be substituted as a whole by a preform such as she.
The constituents of (9) can be isolated by substitution as
well as permutation.
The delimitation process can become increasingly
difficult in cases where we have more than one prepositional phrase and/or a nominal phrase. The question is: Are they instances of independent constituents or members of a complex syntactic constituent, e.g.: (1O)
Sie standen auf der Straße nach Hamburg. 'They were standing on the road to Hamburg. 1
29
Out of five possible permutations only one is acceptable, namely, (lOa)
Auf der Straße nach Hamburg standen sie. On the road to Hamburg they were standing.'
Auf der Straße nach Hamburg is also substitutable only as a whole by a proform of location: (lOb)
Sie standen dort. 'They stood there.'
(We)
*Sie standen dort nach Hamburg. 'They stood there to Hamburg.'
The dependency relation of nach Hamburg in relation to auf der Straße
is
shown by the possibility of eliminating nach Hamburg (lOd)
sie standen auf der Straße.
(lOe)
*Sie standen nach Hamburg.
(1Oe) would be possible if understood in a context like: train and had to stand all (11)
the way to Hamburg. '
'They took the
An English sentence like
They waited on the road to Baltimore.
could be subjected to the same tests as the German exanple (1O).
Evidently,
(1O) and (11) are instances of structural compression, where the attributive component represents the notion road leading to X', the 'directional1 predication is downgraded and dominated by 'location'. (12)
Er kaufte das von seinem Freund vor drei Jahren am Stadtrand erbaute Haus. 'He bought the house that had been built by his friend three years ago in the outskirts of the city.'
Such complex attributive structures would, nevertheless, first be isolated as a whole; this example illustrates that possibilities for structural compression vary from language to language (in kind and degree). Combined prepositional and/or nominal phrases that are substitutable and permutable only as a whole are instances of one (complex) syntactic constituent.
It is of interest to note that the attributive elements can be
30
instances of various semantic concepts (cases), such as location,' direction, object, benefactive (see also 2 . 2 . 4 . 1 ) . In the following example, prepositional phrases are substitutable and permutable individually. (13)
sie saßen am Vormittag im Garten. 'They were sitting in the garden in the morning.'
Any attempt to permute the two prepositional phrases as a whole results in unacceptable sentences (13a)
*Am Vormittag im Garten saßen
sie.
'In the morning in the garden they were sitting. 1 In case of substitution, each PP (prepositional phrase) has its own proform. (13b)
Sie saßen dann dort. 'They were then sitting there."
As a result, both PPs are instances of independent syntactic constituents. As a last possibility we have examples where the two phrases can be substituted and permuted separately as well as together. (14)
Er rasierte sich in der Toilettenanlage im Bahnhof. "He shaved in the bathroom at the railroad station.'
There are doubts as to the validity of this possibility as identification of separate syntactic units, since there seems to be a difference in 'information content1 involved if they are substituted/permuted as a whole as opposed to separately. If substituted by a proform, only one phrase can be covered by this proform. (I4a)
Er rasierte sich dort in der Toilette.
(14b)
*Er rasierte sich dort dort (in ihr?).
This indicates that both phrases are in a close semantic relationship, possibly one of inclusion. Comparing (14a) with (14c)
Er rasierte sich dort im Bahnhof. 'He shaved there in the railroad station.'
31
indicates a 'Pars-Totum' relationship between the two PPs, where it appears intuitively more natural to use a proform for im Bahnhof (Totun) than for in der Toilette (Pars);
(14c) seems to change the "information content" by
eliminating the 'Pars-Totum' relationship,
im Bahnhof can be interpreted
here as an apposition which expresses relational identity between dort and im Bahnhof. That two independent syntactic constituents can be instances of two very closely related semantic concepts is shown by such examples as (15)
Unser Onkel besuchte uns vor zwei Wochen einige Tage. Our uncle visited us for a few days two weeks ago. '
where vor zwei Wochen refers to time (point in time) and einige Tage to time (duration). 2)
Not permutable and not substitutable:
In German, the es functioning as a 'correlate' for subject position is subject to the above restrictions: (16)
Es betraten drei Männer das Lokal. "There were three men entering the pub.'
Es and there are not 'syntactic constituents' in our sense, but rather semantically empty 'correlates' required by the syntax of specific languages, such as English and German. In these instances the actual subject must appear at another position in the sentence (usually immediately following the verb). 3)
Permutable, but not substitutable:
At first, the es (it) in (17)
ES regnet heute. 'It is raining today. 1
appears to be the same as the one above in (16). ever, reveals that this es is permutable: (17 a)
Heute regnet es. 'Today, it is raining.'
A close examination, how-
32
Evidently, this es is not just a "correlate1 but rather a positional proform (Stellungsglied, Brinker 1972:122), that is semantically void.
It can
fill any position the subject could take in the syntax of German or English We are new in a position to distinguish between three different es, namely: 1)
es as 'correlate1 (Platzhalter) -permutable, -substitutable
2)
es as 'positional proform' (Stellungsglied) +permutable, -substitutable
3)
es as 'proform1 for a syntactic constituent (Satzglied) + permutable, + substitutable (Brinker 1972:119-23)
This distinction is important since only es 3) plays a role in valence considerations.
Repeatedly, es 1) and es 2) have been treated as instances of
valence, which resulted in attributing to such verbs as rain the valence 2 factor one on the expression level and zero on the content level. 4)
Not permutable, but substitutable:
This possibility is only of interest in nonfinite verb structures, e.g. (18)
Durchgehen verboten. 'Trespassing prohibited.'
In summary, it can be stated that substitution and permutation are basic criteria for the isolation of syntagmas of different complexity. According to tests 2) and 3) only es 3) is a candidate for valence consideration. 2.2.3
Syntactic constituents as a testing ground for the nuclear/peripheral (obligatory/free) distinction
In Section 2.1.3.2, HuS's position concerning valence was discussed in some detail.
An important criterion in distinguishing prepositional objects
A detailed account of es as a criterion for distinguishing avalent verbs (nullwertige Verben) is given by Heringer (1967:25-31).
33
from adverbial phrases was the substitutability of the preposition involved, so that, for example, in (4) Er wartet auf den Freund, auf is not substitutable and auf den Freund, therefore, is considered a prepositional object. On the other hand, auf is substitutable in Er wartet auf dem Bahnhof and therefore auf dem Bahnhof is an adverbial phrase, i.e. adverbial phrase of second degree. Although HuS claim to base their analysis on syntactic criteria, such as the deletion test, it is evident that they resort to logico-semantic concepts in their tripartite distinction 'obligatory/facultative/free' on the one hand, and to pragmatic concepts in their 'necessary/not necessary' distinction on the other. Götze (1974) makes a good point when he criticizes HuS's mixing of levels.
What syntactic constituents can be omitted depends to a consider-
able extent on the communicative-textual situation. only be considered in a contextual grammar.
These can, however,
When Götze maintains that
spoken language is full of structurally incomplete utterances, and that they constitute the rule rather than the exception, he is right.
He sees, there-
1
fore, no justification in taking the "normal case as the basis in foreign language instruction (Götze 1974:66). We do not agree with this contention on the basis that all elliptic or fragmentary utterances presuppose knowledge of the complete utterance, among other things the actants required by the valence of the verb.
The under-
1
standing of such "incomplete utterances is an instance of a higher level of language competence. 2.2.3.1
Regarding the distinction "obligatory/facultative1 versus 'free 1
As explained earlier, HuS use the deletion test to distinguish between obligatory valence, facultative valence, and free adverbials.
To distinguish
free adjuncts from facultative actants, the adverbial clause test is applied Helbig considers facultative actants as part of the proposition (enge Verbergänzungen) on the 'deep structure level', and thus they fall within On the possibilities for elliptic structures and their effect on the remaining actants, see also Kolvenbach ( 1 9 7 2 : 9 1 ) .
34
the valence of the verb.
Facultative valence, however, he considers a
matter of surface structure.
Not only does Helbig resort to different
levels, and it becomes uncertain whether his "syntactic valence1 is really syntactically defined; he also makes the distinction between parts being 'necessary1 (notwendig) or 'not necessary1 (frei) (1971b:272). This results in a circular argument.
How can one arrive at an adequate
description of valence in Helbig's understanding, if actual utterances with possibly omitted 'facultative actants1 are the basis for description? Syntactic tests as used by HuS are not sufficient to make the obligatory/ facultative/free distinction. 1
'necessary
The circular argument is also revealed when
is further defined as 'not to be eliminated', since elimination
would result in ungrammatical sentences.
At the same time "facultative
1
actants' which are "necessary can sometimes be omitted. As far as the distinction 'verbal complement' versus 'free adjunct' is concerned, it has already been observed that HuS's two tests are not sufficient to arrive at such a distinction.
Another approach in the dis-
1
tinction 'verbal complement versus 'free adjunct' is the consideration of whether the syntactic constituent (syntagma) contributes to the meaning of the sentence (Satzbedeutung) or whether it adds information independent to that required by a certain predication. It can be assumed that the more restricted the occurrence of syntagmas, the closer and more specific is their relation to each other and to the predication under consideration.
It is, for example, not accidental that
the "animate subject/inanimate object' relation accounts for a large group of two-place verbs in many languages.
The formal distributional character-
istics of syntactic constituents are, in many instances, reflections of semantic relations.
It must be stressed that a certain distributional type
can include, paradigmatically, syntagmas of various forms (NP in an oblique case, prepositional phrase, adverbial); i.e., a type is not a group of syntagmas of one specific case or of one specific morpho-syntactic composition.
It is also true that one and the same syntagma can be a member of
different types.
It appears that Gisela Zifonum's types (1972) are rather
35
close to the concept of 'relational constants' (see 2.2.4.3).
The following
may illustrate the various types as suggested by Zifonum:
habreg ηsterb-
der Mann das gro e Haus ich
komm-
arbeit, fahr- ,
dem Jungen an seinen Bruder dieser Tatsache einer Frau
ίη einer Stunde
ί morgen
ί
auf der Stra e hinter dem Haus im Park dort
samstags nach der Schule um 12 Uhr
voller Hoffnung in Freude in gro er Menge
(Based on Zifonum 1972)
The following definitions of Zifonum are interesting in regard to the actant-predicate relation and the predicate-circonstant relation: 1
Valence is defined i.s the ability of a predicate to associate with an n-tuple ( o f ) quantities of syntagmas that are in an interdependency relation with it (i.e. are obligatory).
2
All types where at least one syntagma occurs with one random predicate (p^) within the n-tuple, are Α-types (free adjuncts—'Angaben'); all types where no syntagma occurs with a random predicate ( p . ) within the n-tuple, are Ε-types (necessary actants— 'Erg n z u n g e n ' ) .
3
Types with non-substitutable prepositions are E-types; but not all types with substitutable prepositions are A-types. (My translation of Zifonum 1972:189, 195-97)
In sumnary, all positions required by the 'governing characteristics' of the verb are valence positions; these include subject nominative, genitive, 4 dative, accusative, prepositional phrases, and adverbials. The possibility Helbig wants to draw a clear distinction between "Valenz" and "Rektion", on grounds that "Rektion" is restricted to "objects" while "Valenz" can include prepositional phrases and adverbials (Helbig 1971b: 275) .
36
of surface cases and prepositions having multiple functions must be stressed (see also Nilsen 1972:21), e.g. German accusative (Object, Temporal, etc.). The sane holds true for dative as a position for (19)
Peter hilft
uns.
'Peter helps u s . ' (20)
-type as well as A-type:
(E-type)
Peter singt uns ein Lied. 'Peter sings a song for u s . '
(A-type)
Only in (2O) can we have the alternate form für uns for uns. Should a surface case stand for two semantic functions, we may have instances of homonymy, as in (21)
Ich schreibe meiner Schwester einen
Brief.
interpretable äs (21a)
Ich schreibe einen Brief
an meine Schwester.
write a letter to my sister. 1
(21b)
Ich schreibe für meine Schwester einen
Brief.
write a letter for my sister.'
Heibig considers an meine Schwester a "facultative" actant, i.e. part of the verb valence; and für meine Schwester, a "free adjunct" (dativus coimtxü).
This indicates again the vagueness of the "facultative" concept.
Or must we enter two verbs sehreiben, one with valence^.,. /., ^ , and the other with valence? HuS's entry for schreiben in their dictionary is 'schreiben..+ , 3 > _ 4 ' (HuS 1973:395), corresponding to "someone writing (something to someone about something)', a rather far-fetched interpretation of valence, to say the least. In the area of prepositional phrases, Zifonum indicates the close relationship of German wit-phrases with fee-prefix verbs: beliefern
l mit Ware ) ·! mit Blumen / beschenken [ mit Geschenken )
"supply "present
with goods' I / w i t h flowers' / / with presents')
37
and considers these mit-phrases as E-types based on definition 3) above (see also 3.2.8). At the same time, the possibility of having two prepositional phrases with the same preposition in one sentence is considered an instance of one preposition as part of different types: (22)
Er beschenkte mich mit Freude mit einem Blumenstrau . (Zifonum 1972:196) "He presented me happily with a flower bouquet.'
Mit Freude could be paraphrased by voller Freude.
Mit Freude would be an
Α-type, while mit einem Blumenstrau would be an E-type. Syntagmas, in which the prepositions are substitutable to a limited extent with a certain subclass of verbs, are instances of E-types: komm-
zum Essen
'come
to the meal'
geh-
ins Haus
"go
into the house"
fahr-
nach Mannheim
'travel (go) to Mannheim'
The importance of considering non-substitutable prepositions as carriers of some meaning is evident (with substitutable prepositions this has never been questioned) if we inspect a set such as: bestehen
'exist'
bestehen aus
"consist o f '
bestehen auf
'insist upon'
The prepositional phrase in these examples must be of the Ε-type. Zifonum's types are without doubt semantic and must be seen as an attempt to exemplify differences between Ε-types and Α-types via distributional means, or, in other words, to delimitate those constituents that form the 'semantic nucleus1 of a sentence. Variability is an inherent characteristic of language and the basis for its continuing development.
Any distinction regarding obligatory/faculta-
tive/free valence must take this factor into consideration.
Both Fillmore
(1968, 1971) and Chafe (197O) dealt with the concept nuclear/peripheral, but neither of them leaves us with conclusive criteria to distinguish these controversial concepts (see also 2 . 2 . 4 . 1 . 1 ) .
38
2.2.4
Towards a logico-semantic specification of valence
After this survey of different ideas on valence, from logico-semantic to irorpho-syntactic approaches, we will now attempt to delimit the scope of valence in so far as it is useful for pedagogical purposes. Kelbig's argument about valence centers around the question of whether the concept of valence is to be considered as a formal aspect of the expression level, or whether it is a property on the conceptual level: In Wahrheit werden jedoch nicht alle begrifflichen Relationen in jeder Sprache und erst recht nicht in jeder Sprache in der gleichen Weise realisiert. (Helbig 1971b:279)
It is exactly this difference in realization of 'basic semantic sentence structures' that we are primarily interested in, since it is expected that these 'semantic sentence structures' constitute an acceptable basis for cross-language cotrparisons. The usefulness of 'deep-case-like1 notions was the main topic at a "scientific workshop1 (Wissenschaftliche Arbeitstagung) in Leipzig, 1974. At this workshop Helbig still insisted on the importance of distinguishing various levels of valence, but, as reported by Perl, concludes that Fillmore's case grammar should be incorporated into further work in valence theory: Jedes Valenzmodell hat seine Ziele, die beim Vergleich unbedingt berücksichtigt werden müssen. So sollte z.B. die Kasusrollentheorie von Ch.J. Fillmore zur Erweiterung unbedingt mit einbezogen werden. (Perl 1974:242)
Heibig (1976) emphasizes the complementary character of models based on syntactic valence, logico-semantic valence and componential analysis. A semantic basis for the description of sentence models appears to be the primary concern of this article. In a lecture at the LSA Summer Meeting 1974, in Amherst, Mass., Fillmore mentioned Helbig's valence dictionary for German verbs and stressed its usefulness. The dictionary is based on difficulties in learning, understanding or perceiving valence and distribution which the foreigner encounters in learning German. Fillmore pointed out that initially his contributions to case grammar were made with a valence dictionary for English in mind. Later on, however, he came to realize that such a goal could not be achieved for English and thus had abandoned this endeavor. Certainly, no valence based on surface verbs completely explains the structure of clauses containing those verbs. To illustrate the point Fillmore gave the following examples: She wore a green dress to the party. She went to the party wearing a green dress. A person's knowledge of meaning cannot be exhaustively described as the knowledge of the meaning of words. We must go beyond words. (Fillmore 1974 lecture) It seems Fillmore's main concern was to show the limits of case grammar.
39
2.2.4.1
Case gramnar in the context of linguistic valence
For the last several years the author of this thesis has followed with great interest various developments in 'case granmar' and its relation to valence theory.
It is certainly not the purpose of this thesis to trace the develop-
ment and various modifications of Fillmore's case granmar model, nor the criticisms that have been raised against it. found in the linguistic literature.
Numerous such attempts can be
"Cases' are units that stand as seman-
tic functions (roles) in relation to the predicator.
The list of such cases
used by Fillmore has varied from 1968 to 1971, and he makes no claim as to the completeness of his 1971 case list in number and type. 1
indicated that there should be a natural "stopping point
It is, however, in assigning
semantic functions to noun phrases occurring in specific syntactic positions (FilliTDre 1971:4O). Where exactly is the 'stopping point'?
In order to consider case grammar
a viable tool for describing semantic-syntactic relations in a non-languagespecified (außereinzelsprachlich) context, the number of these relations must be kept relatively small. The following cases are considered in Fillmore (1971) and are defined by him as follows: Agent (A) Expe r ienc er
Instrument
the principal cause, instigator of the action. (E)
(I)
the entity that undergoes the psychological event or is in the mental state described by the verb. the immediate cause of an event; the stimulus or thing reacted to in psychological verbs; the natural force responsible for the event.
Object (O)
the entity that moves or undergoes change; the 'content' of a psychological event. This remains a 'wastebasket' case for other roles that are not differentiated semantically.
Source (S)
the earlier location, state, or time point.
Goal (G)
the later state, or end-result of some action or change; the later location or time point; in case of an embedded sentence, the resulting state or event in a causative construction.
Detailed accounts are given, e.g. in Huddleston (197O), Abraham (1971:185-205), Cook ( 1 9 7 1 ) , Mellema ( 1 9 7 4 ) .
40
Benefactive (B)
the entity that is the beneficiary of an action. Fillmore (1971:52) indicates that benefactive constructions occur in sentences with agents, and he reconsiders the semantics of sentences with benefactives by introducing a higher sentence analysis.
Location (L)
spatial orientation of action, event or situation.
Path (P)
'itinerative' additional complement type to source and goal.
Time (T)
temporal orientation of action.
These cases, in the order listed above, correspond to Fillmore's hierarchy for subject selection in English: The case in a given sentence which occurs first on this list determines what is to be the subject of the sentence in ... the 'unmarked 1 instance. (Fillmore 1971:42)
This means that the case frame of verbs as they are entered in the lexicon is written so that the cases are in a left-to-right order regarding the surface subject selection hierarchy, e.g. break +
Possibilities for subject selection of "deep cases' vary from language to language. A comparison of hierarchy structures across languages provides valuable insights into semantically-based constraints in the surface realization of cases and, as we shall observe in Section 3.2.2.4, permit certain typological statements on subclasses of verbs. 2.2.4.1.1 Criteria for case identification Being a syntacticist at heart, Fillmore attempts to identify 'role-types' via syntactic criteria, but considers the assumptions: 1)
One-instance-per-clause principle
2)
Natural 'stopping point' in classifying semantic functions
3)
'Emically1 distinct roles
4)
Earlier-later indication as different instances of the sane cases
41
... fairly vague, they seldom lead to beautifully unambiguous results, and they are always subject to other sorts of considerations. (Fillmore 1971:42)
Fillmore's assumptions when compared to the syntactic tests of Helbig (2.1.3.2) and Engel (2.1.3.3) fail to specify case roles any more explicitly than HelMg and Engel identify obligatory actants. tions 1 and 4 are contradictory.
Incidentally, assump-
Fillmore's observation that he no longer
confuses relational notions with true case-like notions (Fillmore 1971:42) still does not provide us with criteria to decide what these "case-like notions' really are. 2.2.4.1.2 Nuclear versus peripheral cases One of the major difficulties in case assignment is the question of whether a certain case is nuclear (prepositional) or peripheral (modal). observed a similar difficulty in valence theory (see 2.1.3.2).
We have Fillmore, in
1971, seems inclined to introduce modal elements, such as time and place, from higher predicates.
It is, then, such a higher verb (occur,
which takes location and time cases.
happen)
He observes, however, that there are
verbs which take location and time complements directly and, therefore, these are nuclear (Fillmore 1971:49, 50).
We are not provided with explicit
criteria on how to differentiate between nuclear and peripheral cases. In his lecture (1974), Fillmore pointed out that the study of the valence factor of lexical items is not sufficient for a theory of sentence comprehension. He emphasized the importance of the rhetorical organization of a text for the interpretation of the sentence and referred to such areas as emphasis, topic-coirinent, focus, presuppositions, information structure, functional sentence perspective, cognitive meaning versus other sorts of meaning, and scope.
The only viable representation of case structures which Fillmore considers is a dependency notation making use of "stemmas". The semantic interpretation of the sentence is a collection of "stemmas" plus information on variable relation. These suggestive notions are not presented in any formalized manner.
42
In summarizing our observations on Fillnore's case gramrar, the following is important in the context of this thesis: 1)
We shall consider for incorporation in our framework of valence description such general case notions as Agent, Experiencer, Instrument, Object, Benef active, Location and Time.
2)
Fillmore's criteria of 'contrast' and 'complementarity' are not considered useful for the identification of obligatory actants; instead, the tests discussed in 2.1.3.2, 2.1.3.3 and 2.2.3 will be enployed.
3)
Due to the pedagogical orientation of our considerations and the two-level specification intended, relational constants are presented in the order they would occur in an actual, affirmative, non-contrastive sentence.
2.2.4.2 Semantic sentence structures as representations of cognitive concepts In searching for a certain formalism in semantics (see also 2.2.1), an at-·· tempt is made to arrive at general principles that underlie the organization of human knowledge. It is especially this aspect that appears to us of utmost importance if we want to suggest dependency (valence) structures a that are psychologically acceptable as well as universal enough to permit cross-language comparisons. Not all of our thoughts can easily be converted into linguistic form. The meaning of various expressive forms of art, for example, cannot be readily put into words; neither is there a need to do so, since the receiver or observer has at his disposal different means for interpretation (understanding) besides linguistic ones.
Universal in our context is not to be interpreted as present in all languages, but rather in more than one language; even the Universalienprojekt is working with this limited concept of universality (Seiler 1973:14)..
43
We will not concern ourselves with these other aspects of expression and interpretation of thought, interesting and challenging though this might be. Rather, we will concentrate on semantic structures that act as a filter " through which thoughts trust pass before they can eventually be converted into sound" (Chafe 1971b:57). The multidimensionality of conceptual space requires a variety of 'semantic constraints' applied to thoughts as they are brought into the linguistic domain. That these constraints vary, not only from person to person, and from one social group to another, but also from language to language, is important for our understandijng of the differences in knowledge of the world. Again, our interest can only be in those constraints that are cotmonly accepted by a large population within a particular language (or dialect). It appears that certain cases (relational constants) share characteristics. As an example, Agent and Instrument have some component of causality in common, while Cfoject and Benef active seem to share a notion of noncausality. On the assumption that relational constants both constitute linguistic universals and represent conceptual universals, they are selected as a basis for our investigation (see also Kintsch, 1974). 2.2.4.3 Features as criteria for the identification of relational constants Semantic constraints within a 'semantic sentence structure1 are regulated by features (Chafe's inherent features and derivational features). Let us first consider the notion of causality. Not enough attention has been given to the distinction of different cases that could all be subsumed under the notion 'causality1. 'Cause' is here considered in a wide sense. Without any doubt the most prominent candidate in regard to causality is the Agent (A), as specified by Fillmore (1971). In this thesis we follow Huddleston's suggestion to introduce the feature Intent (197O:5O6), with only the modi l ι +Intentl, to account for sentences such as:
I
(23)
Ich habe das Brot geschnitten. Ί cut the bread.'
(24)
Ulntent
Ich habe mich geschnitten. Ί cut myself.'
-Intent
(most probably)
44
So far, then, we have for Agent the specification
+Intent .
The Intent
feature gets a definite negative reading in the Instrument (I), which would be specified as [-Intent] . An interesting difference between the much discussed WITH/USE TO pair comes to light when seen in the context of the above features (Nilsen 1973a: 91): (25)
Ich zerbrach die Vase mit einem Stuhl. broke the vase with a chair. 1
(26)
[j^Intent]
Ich verwendete einen Stuhl, um die Vase zu zerbrechen. 'I used a chair to break the vase.'
f+lntent]
Without this specification, (26), A (Agent) can be ambiguous with regard to the Intent feature; this is not the case with I (Instrument). Further observations with the cause-notion reveal sentences that do not fit within the above A/I distribution, e.g.: (27)
The wind opened the door. 'Der Wind öffnete die Tür. 1
(28)
The hurricane destroyed the village. 'Der Hurrikan zerstörte das D o r f . 1
(29)
The falling branch damaged the car. 'Der herunterfallende Ast beschädigte das Auto.'
In the above sentences we cannot consider 'the wind1, 'the hurricane1, or 'the falling of an Agent. meationed by we introduce
branch" as instances of Instruments that are under the control We set up a separate relational constant: Force (F), (also Huddleston 1970:503). To distinguish Force from Instrument, the feature [+Control| ; attributing the f-tControl] to Force
and the [-Control] to Instrument. causative relational constants: Agent
f+Controll f+Intent 1
Force
So far, we distinguish the following
f+Control [-Intent 1
Instrument [-Control] [-Intent]
45
Incidentally, the nonsiraultaneous occurrence of A and F in the sane proposition speak in favor of considering them as separate relational constants. Although the following could be an idiosyncracy of German, it is striking that the three relational constants above have distinct prepositions associated with them, e.g.: (30)
Die Fensterscheibe wurde von mir zerbrochen
'The window was broken by m e . ' (31)
Die Ratten wurden mit Gift
A
getötet.
'The rats were killed with poison. 1 (32)
I
Das Dorf wurde durch einen Hurrikan zerstört. 'The village was destroyed by a hurricane."
F
In order to consider sentences with a 'higher-cause1 predicate, we add another causative relational constant. For lack of a better name, we shall call it Cause (C); it corresponds to the notion (Bikausal-Funktion), as discussed by Heger (1971), and stands for the concept that a cause outside the 'instigator-causer1 (Agent) acts upon the Agent to make him do something. An instance of this relation would be: (33)
1· made him wash my car.
'Ich hieß ihn meinen Wagen waschen." This concept is present, even if the 'down-graded1 agent is not present: (34)
I had my car washed.
'Ich ließ meinen Wagen waschen.' Notice the difference in English between 'make' and 'have + past participle1; and the anbiguity resulting from using 'had* in (35)
I had him wash my car.
Applying the features C
-Ktontrol l+Intent |
^Control A
and +Intent , we have
+Control [+Intent j
F
Ucontroll -Intent |
I
-Control -Intent
46
With the above specification, we arrive at a + activity scale, with Cause (C) leading on the plus side and Instrument (I) on the minus side (see also Nilsen 1972:49).
The relevance of ranking according to a + activity
in matters such as primary topicalization will be put to a test when comparing systematic differences in primary topicalization between languages, such as English and German. q
A content on the relational constant Force as in (36)
Darkness frightened
the child.
"Die Dunkelheit machte dem Kind Angst.'
is necessary.
Here the relational function between 'darkness' and 'frighten'
is an 'experience-causing1 one.
Even in this interpretation, the
+Control]
[-Intent) features seem not to violate the actual process. There is no need to resort to such features as 'animateness', which, if used at all,
should be reserved for the specification of the participant
(actant) itself.
There are, however, obvious connections between such
relational features as 1
+Intent
and 'animateness'.
Nilsen uses the fea-
1
tures 'animate , "count , and 'concrete' as 'semantic features for case distinction' (Nilsen 1973:112-120).
His major concern with the instrumental
case in English causes him to subdivide the instrumental case into: Material-case, Force-case, Bodypart-case, and Tool-case (Nilsen 1973:155ff). Since we have already discussed and specified the relational constant Force, it is excluded from the 'instrumentality' range.
Instrument is
often referred to as an instance of secondary cause, but when the A(gent) does not occur the I(nstrument) is no longer a secondary cause and may be treated as an agent. (Cook 1973:59)
(37)
John broke the window (with a hammer). \+
AO
'John zerbrach die Fensterscheibe (mit einem Hammer).' (38)
The hammer broke the window.
Α Ο
*'Der Hammer zerbrach die Fensterscheibe."
Some arguments in regard to the Agent/Force distinction are set forth by Fillmore without arriving at conclusive answers (Fillmore 1971:74).
47
In German, the topicalization of Instrument is problematic for most verbs.
Sentence (38) in the passive would be: (38a)
The window was broken by the hummer. *'Die Fensterscheibe wurde von einem Hammer zerbrochen.'
Even if the alternate choice is considered, of taking 'hartmer'—'broke 1 as a 'Force' relation, the German sentence above is not possible. (38b)
(*)Das Fenster wurde durch einun Hairnnvr zerbrochen.
Therefore, once the Instrument relation is established (and this is a semantic relation) it cannot: be changed to an Agent relation without changing the meaning of the sentence.
The fact that 'hammer' may appear as the subject
in English does not change its semantic relational aspect, that of being Instrument.
The categorical relatedness of active and passive have long
been established.
Their difference lies in 'pattern-meaning'.
The following remarks on the Instrument case relation must be seen in their potential for pedagogical considerations of case incorporation (lexicalization).
Following Brekle (1970:72, 81, 8 2 ) , we observe a close rela-
tionship between aff(ective)/eff(ective)
and the semantic marking of I;
ultimately this difference is caused by the relational constant dominating the verb.
The aff-relation corresponds in case grammar to the 'most neutral
Objective case.
Even if no difference can be found on the morpho-syntactic
level between aff and eff, we want to suggest these as subtypes of the relational constant Ο (Object) providing we can posit a sufficiently general difference ΙΛ their semantic relation.
It might very well be that in dif-
ferent languages topicalization rules differ according to this distinction. At any rate, it can be observed that the Instrument relation, as added to the 'semantic sentence structure' in (39) and (4O) below, reveals different kinds of instrumentality. posive (intentional) sense
Both 'with a knife' and 'with flour 1 have a purhowever, while 'with a knife1 stands in combina-
tion with aff, 'with flour' is under eff.
This results in an interesting
distinction regarding the instrument, namely: 1)
If the respective 'verb' is dominated by eff, then the INSTR automatically is semantically marked as 'stuff' (material).
48
2)
If the respective 'verb' is doninated by aff , then the INSTR automatically is semantically marked as 'tool1 (device) .
We refer to relations like the following: (39)
bake a cake with flour Ο ... *—»I„ J (Material/Stuf f Instrument) eff Mat
(40)
cutting bread with a
knife
Ο ,_, ·«—> I . (Tool Instrument) aff Tool
The Ό ff ·*-* Irryyvi ' relation can undergo the 'do to1 test ; the Ό _-· ' cannot. Various groups of verbs, such as 'fill-in1 verbs, verbs indicating change of appearance, taste (sugar, dye, . . . ) , and verbs of enclosure (screen, board, ...) permit or are instances of Material Instrument incorporation. (41)
John filled
the turkey with
stuffing.
*'John f llte den Truthahn mit F llung. 1 (41a)
John stuffed
the turkey.
'John f llte den Truthahn.' (42) (42a)
The workmen covered the tennis court with asphal t . 'The workman asphalted the tennis court. 1 (Nilsen 1973:155-56)
The same holds true for the 'Tool Instrument1 . In order to keep the 'information content1 of the following English sentence the same as in German, the 'Tool Instrument1 must be expressed. (43)
John taped a Picasso print to his bedroom ceiling. (Nilsen 1973:165)
(43a)
1O
John befestigte mit einem Klebestreifen einen Picasso-Druck an seiner Schlafzimmerdecke.
Applicability of the 'do to1 test results in a 'change of state' in the object as a result of the action. This 'change of state* is interpretable in a wide sense (Anderson 197Ib).
49
Or, another instance: (44)
He killed her with a knife. 'Er tötete sie mit einem Messer. 1
(45)
He knifed
her *(with a
knife).
'Er erstach sie (mit einem Messer). 1
While the English verb 'knife' has the 'Tool Instrument' incorporated, the German erstechen which partially corresponds to stab, does not inply the specific tool used in the action. Consideration of a 'Body-part Instrument1, finally, could prove extremely productive in vocabulary expansion. Let us just consider verbs of motion in which various 'body parts' are incorporated, such as nod (head), soon (eye), grasp (hand), kiss (lips). Incorporated Instruments add a particular meaning to the verb and therefore are barred from surface realization with that particular verb. However, if specified, the Instrument can appear in the actual sentence: (46)
Jürgen küßte Ulrike *(mit den Lippen). 'Jürgen kissed Ulrike *(with his lips).'
(47)
Jürgen küßte Ulrike mit schokoladebeschmierten Lippen. 'Jürgen kissed Ulrike with chocolate-smeared lips.'
The fact that case incorporation can differ across languages motivated us to make these further distinctions within the relational constant Instrument. In surmBry, the group of relational constants considered so far on a + activity scale are: Body part Material/Stuff T001
W
50
2.2.4.4
The non-causative relational constants
Chafe's (1970) major concern for the well-formedness of semantic structures supports our contention on the primacy of semantic well-formedness over syntactic well-formedness.
This is not to deny the importance of morpho-syn-
tactic forms as they reflect semantic concepts; there is, however, considerable redundancy on the morpho-syntactic level, especially in a language such as German.
We accept Chafe's basic verb classification, according to
intrinsic features, into:
state, process, action, and action-process verbs,
on the basis of their universal validity (see also footnote 8) and their easy applicability in a pedagogical grarmor.
In considering further case
relations, we make use of these basic verb types. 2.2.4.4.1
The Object relation
Let us first consider the 'wastebasket' case, the Object' (Fillmore), which corresponds to "Patient" (Chafe), since it represents the notionally nost neutral case relation. (48)
John is dead. 'John ist tot."
(49)
°s(tate)
The diess is tight. "Das Kleid ist eng.'
(50)
Ο
The vase is broken. "Die Vase ist
zerbrochen."
O
s
Without entering into the discussion about the "central" status of the adjective, we do include adjectives in our semantically-based valence concept. There are, after all,
languages (Navaho) where the copula is not necessary
as a syntactic link (Anderson 1971b:38, note 1). In his relational clause, Halliday (1971) distinguishes between "attributive" relation and 'equative1 relation. One distinction between the two is that the 'equative1 is reversible and that it has the role 'identifier', which expresses class-identity.
Halliday includes in this type, verbs such
as be, equal, represent, resemble, stand for (Halliday 1971:155).
We will
not consider resemble in this type, since it does not imply 'identity1 but rather sharing some characteristic with the entity compared with. The
51
'attributive', on the other hand, is nonreversible; and here we shall consider instances of total class inclusion (0T
, ) , partial class inclusion
or class membership (Oirajcs /, and resemblance (OKes ) . D An interesting observation made by Halliday with regard to relational clauses is that in relational clauses, quite unlike in clauses of action or of mental processes, the verb is regularly unstressed.
This is a symp-
tom of the much weaker function of the verb in the clause (Halliday 1971: 155).
The same phenomenon can be observed in 'function verb constructs' of
the type to bring to a conclusion (zurr. Abschluß bringen) . The following semantic functions of the Object case relation will be distinguished in this thesis:
O. Id (entificatiori) (51)
Dieses Mädchen ist
Erika.
'This girl is Erika. where Erika is the O
, Id
1
(reversible)
Id'
Part(ial inclusion) (52)
Sein Vater ist Arzt.
O 0 Part s
'His father is a doctor.' where: his father
represents 0
.
(non-reversible)
O.
Incl(usion, total)
(53)
Catholics are Christians.
. Incl
s
"Katholiken sind Christen.' where Catholics is O
(non-reversible)
Incl'
:S (emble) (54)
John resembles his
father.
"John ähnelt seinem V a t e r . where his father
is O
Res'
„ Res
s
1
(non-reversible)
52
lnal(ienable possession) (55)
The dog has four legs. 1
Ο
s
Ο
Der Hund hat vier Beine.'
where four legs is Ο
JLTlclX
.
(reversible)
ο Meas(ure) (56)
John weighs 3OO pounds.
Ο
Ο
S ATeas
'John wiegt 3OO Pfund.' where 3OO pounds is O „ Meas This relation occurs with other verbs such as cost, measure, read, etc. A subdivision into 'Measureequal 1 and 'Measure-comparative 1 could be made.
If 0 is related to an action verb, as in: (57)
The vase broke.
Ο
'Die Vase zerbrach.'
we have the action-object type. If instead of 0 we have A, as in (58)
He is singing.
A
"Er singt.'
we have the action-agent type. If 0 is related to an action-process verb, as in (59)
He broke the vase.
AO
'Er zerbrach die Vase. 1
we have the action-process/agent type; however, since this implies the causative notion, not only A but also F could occur. With O, A, and F we could then have the following case relations with basic verb-types;
11
In the schematic presentation we follow Cook's matrix (1972) in its modified version (Zarechnak 1974).
53
Basic Verb Type
2.2.4.4.2
Ο
A/F
state
Ο
process
Ο
action
A
actionprocess
A/F
S
0
The Experiencer relation
The basic definition of Experiencer as expressed by Fillmore (1971) and Chafe (1970), and stated in 2.2.4.1, is accepted here.
It is sometimes
difficult to draw a clear line between 'actions' and 'mental processes'. For 'mental process" clauses Halliday considers "inherent roles' such as "processor" and 'phenomenon'—in case gramer terms, Experiencer and Object. Halliday1s reference to the nonlimited character of the 'phenomenon' in 'mental process' clauses is worth mentioning: They are distinct in that the 'phenomenon 1 —that which is perceived, reacted to, etc.—is not limited, as are the participants in action clauses, to the class of "things', namely persons, objects, abstractions and the rest of the phenomena on the plane of experience. (Halliday 1971:153)
Considering the Experiencer with basic verb-types, we have the following possibilities: Basic Verb Type
12
A/F
E
state
E
process
E
action
A/F
E
actionprocess
A/F
E
0 s
O
It must be stressed that the author is fully aware of the vagueness
of these definitions and the criticism of this vagueness (e.g. Finke 1974).
54
Mental states (events) can be instances of such notions as perception, reaction, cognition, or verbalization.
Some examples may illustrate the con-
cept of Experiencer: (bO)
Tho child was afraid
E
s
'Das Kind hatte A n g s t . " (61)
The child got frightened.
E
"Das Kind bekam A n g s t . " (62)
Kurt frightened
the child.
AE
"Kurt machte dem Kind A n g s t . " (63)
The noise frightened me.
F E
'Der L rm machte; mir A n g s t . "
The possible ambiguity of (62) results from the [^Intent] feature of the Agent relation.
The German rendering with Arvjr.t machen (cause fear) makes
the above ambiguity rather improbable.
Fillmore suggests the inclusion of
Instrument for the description of psychological event predicators to handle such sentences as (64)
The noiso reminded mo of the accident. Ι
Ε
Ο
(Fillmore 1971:53)
We suggest that 'noise1 is an instance of the "Force" case relation as introduced above, and that Instrument is not needed, although it can occur together with Agent, e.g.: (65)
He frightened A
as a facultative element.
the child with his loud voice. E
l
Since we are interested in including more complex
clause structures, Halliday's reference to 'metaphenomenon1 ('fact 1 and 'report') deserves attention (1971:154). (66)
I noticed lier new arrss.
Ε Ο
' I c h bemerkte ihr neues K l e i d . ' (67)
It. worries me that you look so tired. Ύ metaphenomenon ( f a c t ) 'Es macht mir Sorge, da aussiehst.'
du so m de
Ε Ο
55
There are significant differences in the topicalization of State/Experiencer between German and English.
While in English, State /Experiencer is
readily available to be a subject, in German it is frequently realized as an object (grammatically) in such sentences as: (68)
I do not fed Ε
well
Ο
'Mir ist nicht gut. '
E
(69)
O
I am cold. E
O
'Mir ist kalt. '
The potential for errors that this factor provides is evident. The usefulness of distinguishing between State/Experiencer and Process/Experiencer can be observed in such examples as: (70)
(71)
He is healthy.
Ε
'Er ist gesund. '
(objective)
He feels healthy.
E O
S
Ο
(but not necessarily is! ) 'Er f hlt sich gesund. 1
(subjective)
2.2.4.4.3 The Benef active relation Again we use the definition as outlined in 2.2.4.1, even though Fillmore points to some "unhappy facts about Benef active constructions' (Fillmore 1971:52) . In the framework of basic verb types we can distinguish the following possibilities.
56
Basic Verb Type
A/F
B
state
B
process
B
action
A
B
actionprocess
A
B
0 O
S
O
0
B = DIR /DIR
Since such verbs as sell, buy, give, etc. indicate the transfer (change of possession) of 'something1, the concept of DIR(ection) is a useful one.
It permits the Source/Goal distinction which accounts for the
main difference between such pairs as buy/sell or borrow/lend, give/receive. In the above matrix, the Action-Process/Benefactive verbs would then have the B-relation specified as DIRg/DIR^
This DIR is, of course, distinct
1
from 'direction in the locative sense (^Q·^ · In German, it can be realized morpho-syntactically by the dative case. (72)
Ich gebe (hänge) das Bild an die A
Ö
L
Wand. . Dir
'I put the picture on the wall." (73)
ich gebe dem Vater das Bild. DIR,
0
"I give father the picture.'
Obviously the geben in (72) has a meaning different from the one in (73) due to their different "semantic nuclei" (Satzbegriff).
Dative case in German
can fulfill various functions; in (73) it reflects the DIR(, and is obligatory.
DIR can appear as subject nominative in the passive in English; but
in German it must remain in the dative case. (74)
In the sentence:
Ich hänge dem Vater das Bild an die
Hand.
dem Vater is FIN (dativus commcdi), free, and not part of the prepositional core.
Dem Vater could be paraphrased with für den Vater
2.2.3.1).
(see also
57
FIN is defined as a complex "semantic sentence structure1 and can be realized norpho-syntactically in different ways.
Observations of this kind
(Ιγ.. /FIN/DIR) are extremely relevant in foreign language teaching if we desire to go beyond the 'sentence model" concept.
It is in this context
that the multiple function of the oblique surface cases and prepositions can be described and understood (cognitive learning). 2.2.4.4.4 The Locative relation The distinction L
(stative locative) from
.
(direction) is significant.
One reason is that L can be obligatory as well as outside the proposition, fa
while L-,·- seems to appear only within the prepositional core.
Basic Verb Type
A/F
state
0
process
Ο
action
A
actionproc€!ss
A
L
0
L
s
L
L
Ο
L
In German we observe a differentiation between LS and L_. in the morphoL/IJT syntactic realization of the noun phrase following a specific group of prepositions, such as an, auf, hinter,
ber, etc.
The difference is indicated
not by the preposition but by the case (Ls = dative, L». uii = accusative). In German as well as in English, the Locative (LS) can be realized, in a restrictive sense, by the subject nominative, e.g.: (75)
It Is windy in Montreal.
(76)
Montreal is
OL
windy.
We say "restricted1 because only with certain prepositions such as the prepositions in and through, and only with LS, is SUch a topicalization possible. Comparing such examples as (75) and (76) above with meteorological verbs such as vain, we observe: (77)
It is raining in Washington.
(78)
*Hashington is raining.
58
which is another proof of the restricted possibility of topicalizing L. For further specification of L, the Source/Goal as well as Path concept could be used. 2.2.4.4.5 The Time relation Even though the number of verbs that require a relational constant 'Time' in their 'semantic nucleus' (at least in such languages as English and German) are limited, we have, nevertheless, included Time in our valence study.
Zarechnak refers to a number of meteorological verbs that "require
mandatorily a filler referring to time' (1974:41), e.g. develop, forecast, anticipate, follow, etc.
A comparison of occur with the German es gibt in
the temporal existential sense clearly supports Zarechnak's notion of not treating these verbs as simple statives (O ) .
In making use of the last
available column in the revised matrix, we could include such time-related verbs in our valence specification. Basic Verb Type state
Ambient Ο
Met
Τ
(Meteorological)
process action
Ο
Τ
actionprocess
To consider take place as an instance of an action/time verb seems to agree with the intuitive judgment of a native speaker; it is, however, evident that such other concepts like Location also show close affinity. The same holds true for the German stattfinden. (79) (80)
As to Lyons' examples:
The demonstration took place on Sunday, demonstrated on Sunday. (Lyons 1968:347)
59
it appears that—leaving any constituent structure argunent aside—while take place focuses on the event the demonstration and excludes the agent, demonstrate focuses on the agent, as can be seen if the temporal element is omitted (due to context): (79a)
The demonstration took place.
Ο (Τ)
(8Oa)
They demonstrated.
A
Take place refers to a certain point in time (Tpunct)· In the sentences: (81)
The concert took place last night.
(81a)
*The concert took place for three hours.
The verb last suggests 'duration of time1 ( Τ ρ ) , or 'length of time from beginning to end-point'. (82)
The concert lasted three hours.
(83)
The concert lasted from eight to ten.
(84)
*The concert lasted last night at eight o'clock.
Subgroups within time could be established as follows: Τ Source Τ
Goal
(duration)
2.2.4.5 Summary of relational constants selected for consideration in a pedagogical grammar The primary purpose in our consideration of various models which present syntactic sentence patterns in German was to see how these syntactic patterns reflect semantic sentence structures. Our findings include the following list of relational constants to be considered as a model-like linguistic input for a pedagogical grammar for the purpose of differentiating basic semantic sentence structures in German and English, and possibly other languages.
We make no claim regarding
the completeness of such a list and its applicability across languages.
60
But we do claim that these relational constants are a viable tool for the description of basic sentence structures on semantic grounds and, when supplemented with morpho-syntactic realizations in particular languages, constitute the framework of a linguistic input in a pedagogical grammar. Among the causative relational constants we distinguish the following: C (Cause)
f+control] f+intentl '
A (Agent)
[+control] +intent
F (Force)
[+control] |-intent|
I (Instrument)
-control [ -intent-
-""Mat
(Material, Stuff)
W
(T001)
The noncausative relational constants are: 0 (Object) OId
(Identification)
OircLtt D
(Partial inclusion)
°Incl
(Inclusion, total)
OD Kes
(Resemble)
OT Ο
,
Meas
(Inalienable possession) (Measure)
E (Experiencer) Β (Benefactive) DIR„
(Direction, Source) (Direction, Goal)
61
L
(Location)
L...
(Directional Location)
T
(Time)
T Punct T Dur
(Point in Time)
FIN
(Duration) '(Final)
62
3.
CONTRIBUTIONS TO A PEDAGOGICAL GRAMMAR—THE APPLICATION OF THEORETICAL CONSIDERATIONS
3.1
The experimental language test in German as a research tool in the investigation of errors in the realization of logico-semantic valence
3.1.1
Objective of the test
Due to our major interest in lexico-semantic valence a test was constructed which was verb-centered.
The verbs used in the test were all taken from an
internationally known and widely used German textbook, Deutsch als Fremdsprache 1 by Korbinian Braun, Lorenz Nieder, and Friedrich Schmöe (1967). This particular text was chosen for the following reasons: 1)
It is monolingual, and not directed towards any homogeneous group of learners in regard to their native language.
2)
Its grammar is based on sentence models as discussed in the theoretical part of this thesis.
3)
The vocabulary has been carefully selected not only on the basis of frequency but also on the basis of communicative significance.
All verbs (predicates) used in Deutsah als Fremdsprache 1 (hereafter referred to as DaF 1) were analyzed according to HuS's valence dictionary. These verbs, approximately 21O in all, formed the basis for the verb-centered test. The objective of the test is not to test the students' proficiency in German, although the appearance of the test may give rise to that impression. Our interest was to obtain student errors that indicated the difficulties with, or lack of knowledge of, the logico-semantic valence of the predicates under consideration.
63
3.1.2
Test design
The overall appearance of the test is such that it stresses the conrnunicative aspects of the language.
The test (see Appendix A) consists of four
parts (1, 2, 3, 4 ) . Part 1 of the test consists of 5O short conversational units of the following type (with the verb entered in the infinitive on the left-side margin throughout parts 1, 2, 3): (Item 7)
(Item lo)
danken:
geben:
A:
Ich möchte
B:
Gern geschehe«.
A:
W
B:
eure Hilfe
hast du das
Telegramm
danken.
?
Lehrerin.
Although the student had to prove his overall conprehension of the comnunicative situation and supply all morpho-syntactically relevant information such as correct verb form, and adjective endings and articles (pronouns, prepositions) in the appropriate cases, our main interest is in errors that can be interpreted in terms of logico-semantic valence. In Part 2, a list of words to be used with a particular verb is given. There are 15 items. The; students are asked to form sentences using at least the vocabulary provided, plus additional material necessary to form a correct German sentence. The words provided are not necessarily in the correct order, e.g.: (Item 11)
erwarten:
Kellner - vorige Woche - Chef - Brief
The student's choice of prepositions and cases as realizations of such notions as 0, L, DIFL· are of prime interest. In Part 3, initial sentences are given. The student is to use the "message1 of this sentence in a new sentence where the communicative perspective is changed; or the same semantic sentence structure can be paraphrased.
Directions such as "relative clause1 (Item 9)
Er schrieb einen Brief Er schrieb
and 'passive' are given.
an seine Mutter.
Mutter
64
For our purposes it was of interest to see if the student understood the initial sentence, how he dealt with changes in sentence perspective, and if he was aware of the particular paraphrase structure that was possible. Part 4 presents the student with 45 English sentences.
It is stressed
in the directions that the purpose of this part is not to test the student's 'translation skill1. (Item 33)
He is advised to focus on the verb.
Do you know who broke the vase?
This was undoubtedly the most challenging part since the student was entirely on his own to decide how to present the same 'message' in German. 3.1.3
Test administration
The test was administered to 33 intermediate students and to 31 advanced students of German in the Georgetown Summer Program in Trier, in 1974. The test was given at the end of an intensive six-week language program. The student population taking the test represented a good cross section of American universities and colleges.
The time alloted for completion of
the test was 90 minutes. 3.1.4
Selection of verbs (predicates) for valence analysis
All items that led to errors made by 2O% or more of the persons tested were selected for further analysis.
This is high enough a percentage to insure
that errors made were not primarily the result of other possible sources, such as lack of overall understanding of context, accidental confusion of lexical items, lack of attention, etc.
The verbs in the test items had
been selected after they had been subjected to syntactic valence analysis according to HuS's dictionary. The following verbs from the different test sections were selected to form the basis for detailed semantic valence analyses: Part 1: Item 2
gehen
'to go'
4
hängen
'to hang'
5
kommen
'to come'
6
wohnen
"to live 1 , 'reside'
7
danken
'to thank'
65
8
dauern
'to last 1
9
fragen
1
to ask'
0O
geben
'to give 1
11
gehören
"to belong to', 'possess 1
18
schenken
"to present', 'give as a present'
20
sprechen
'to speak 1 , 'talk'
21
werden
'to become"
23
abfahren
'to depart'
24
ajbstelien
'to leave', 'park'
25
bekommen
'to receive"
26
brechen
'to break'
27
aussteigen
'to get off
29
abfliegen
'to depart by plane', 'take off
31
begrüßen
'to welcome'
33
kommen
(same as Item 5)
35
liegen/legen
'to l a y ' , 'to lie 1 , 'be located'
37
sitzen/setzen
'to sit 1 , 'to seat'
39
anrufen
"to c a l l ' , 'to telephone"
40
ausziehen
'to take off
44
empfangen
'to receive', 'welcome'
47
gelegt/gelegen
"to be p u t 1 , "to.be located'
49
sitzen/setzen
(same as Item 37)
50
stehen/stelien
'to stand', 'be located", 'to put 1
Part 2: Item 2
schenken
(same as 1.18—Part 1, Item 18)
3 gehen
(same as 1.2)
6
abholen
'to pick up'
7
sich anmelden
"to register", 'report to'
8
bleiben
'to stay 1
10
helfen
12
ansehen
'to look at'
14
geben
(same as 1.1O)
15
stattfinden
'to take place"
1
to help 1
66
Part 3: Results from this section will not be used, since many errors occurred in the section which were due to misunderstanding the initial sentence or the terminology used in the directions.
Excluding these items does not matter,
since only 15 items were in this section and the verbs involved in these items appear in other parts of the test. Part 4: The following are the German equivalents of the English verbs being used in the sentences: 1
antworten
'to answer 1
2
hungrig werden
'to get hungry"
3
brechen
(same as 1.26)
4
bringen (nach Hause)
1
5
sich erinnern an
'to remember 1
7
to take home'
'to question 1 , 'interrogate'
befragen
(same as 1.1O, 2 . 1 4 )
9
geben
10
grüßen
12
helfen
(same as 2.1O)
i3
kaufen
'to buy'
(lassen)
'to give regards to"
18
Platz nehmen
'to take a seat'
19
sich treffen
23
sich
24
sich ansehen
"to take a look at'
27
(sich)
'to excuse (oneself) 1
30
gehören
(same as 1.11)
35
beantworten
'to answer something 1
36
müde werden
'to become tired'
37
zerbrechen
'to break to pieces'
40
besteigen
'to climb
41
verletzen
'to h u r t 1 ,
42
erwarten
'to expect'
43
eine Frage stellen
"to ask a question 1
(begegnen)
(ver)ändern
entschuldigen
1
to meet'
'to change'
(a mountain)' "injure"
67
All verbs (predicates) selected fron the different sections of the test as well as related verbs (predicates) that have been included in the valence analysis are given in alphabetical order in Appendix B.
The predicates to
be discussed in the next chapter constitute in no way all the semantic interpretations of individual lexical items.
Idiomatic use is not considered
in our study, nor are different semantic interpretations due to varying stress and intonation patterns. 3.2
Discussion of selected verbs based on presented theoretical considerations—possible application in a pedagogical grammar
In the discussion of selected verbs (predicates), the following procedures will generally be followed: 1)
Selection of predicates according to their correspondence to logico-semantic valence as specified in 2 . 2 . 4 . 3 and 2.2.4.4; e.g. predicates with the semantic sentence nucleus A O E.
2)
Statement on errors made in the test with verbs under consideration.
3)
Entry of the verb under consideration with HuS's syntactic valence specification.
4)
Furnishing of the logico-semantic valence and comparison with HuS's syntactic specification.
5)
Presentation of a two-level specification of the sentence nucleus, with HuS's terminology being used as the morphosyntactic specification of the logico-semantic description in terms of relational constants.
6)
Discussion of the differences in the logico-semantic as well as in the morpho-syntactic specification between German and English.
7)
Consideration of semantically-related verbs in a word-field-like approach.
8)
The discussion of the selected group of verbs is then concluded with a sunmarizing statement indicating the possible difficulties a learner may have with the verbs discussed.
68 3.2.1 On the Agent-Object-Experiencer relation 3.2.1.1 antworten/'fragen 'to answer'/'to ask' Verbs of saying and discourse: (test item)
'to answer 1
antworten
4.1
beantworten
4.35
'to answer something 1
fragen
1.9
'to ask 1
befragen
4.7
'to question (interrogate) 1
eine Frage stellen
4.43
'to ask a question'
Test items with answer and correct responses: 4.1
"Bitte, antworten Sie mir! 1
Please, answer me.
('Bitte geben Sie mir eine Antwort.') 4.35
Have you answered his letter yet?
'Haben Sie seinen Brief schon beantwortet? 1
In 4.1, me was rendered as: mioh (18 students) ; auf mich (1) ; auf mir (1) , beantworte mich (1), beantworte mir ( 1 ) . In 4.35, where beantworten was to be used,geantwortet was given by 37 students. Mistakes made with adjective endings or possessive endings were tabulated, but will only be used when indicative of wrong understanding of the semantic sentence nucleus. The entry in HuS for antworten is as follows: HuS
antworten. ,,, . l+(3)=4
Sn (Sd)
(NS, _ ) daß
(pS)
In a footnote HuS indicate that when the prepositions mit or durch are used, the affinity to the verb is looser; and this Instrument is considered a 'free adjunct1. The three facultative actants specified in HuS are symptomatic of the vagueness of the notion 'facultative'. HuS
beanworten
,n_?
Sn Sa (Sd)
Undoubtedly HuS's descriptions express the essential difference on the morphosyntactic level between antworten and beantworten; on level three some semantic information is supplied. In a footnote (1O7) , HuS mentions that Sj (Dative) is the case of the receiver ('Dativ der sinngebenden Person1 of Brinkmann), and not the person for whom the answering is being done. The correspondence of pS in antworten to S,a in beantworten is also referred to
69
in the sane footnote.
It seems that sane of this footnote information is of
crucial importance; in fact, it is the basis for the valence interpretation considered in this study. Ihe following specification would be given to the two verbs above: antworten
A
(E)
(O)
'answer 1
beantworten
A
O
(E)
'answer 1
It can immediately be observed that antworten only requires an A, while beantworten needs an 0 in addition.
This distinction is not made in English.
As manifold as the sources for mistakes that were stated above may be, there is no doubt that the different obligatory actants and their morphosyntactic realization were ill-understood.
Explanation of the morpho-syn-
tactic behavior of these verbs is, of course, possible without recourse to antworten is a "dative1 verb and beantworten re-
any deeper level; i.e.,
quires an 'accusative object1.
This information, however, does not specify
what these surface cases stand for on the level of the semantic sentence nucleus: antworten
A
(E)
beantworten
A
O
(O),. ^answer (E)·^*^
A
(O)
(E)
These specifications could already eliminate such mistakes as: *Ich antworte die
Frage.
*Ich beantworte. *Hast du seinen Brief
schon geantwortet?
(Test Item 4.35)
The frequency of this type of mistake is extremely high.
These mistakes
are considerably more serious than a missing adjective ending (where the surface case has already been decided by the article and possibly by the ending of the noun).
Such notions as A and E make HuS's information on
level three, such as Sn+Human/+Abstract(as Human), unnecessary, since A and E include this semantic information. enough.
The crucial question is:
But the specification above is not
How can information given by HuS and
notions discussed in the theoretical section of this thesis be ccrrbined
70
to arrive at a satisfactory description of the semantic sentence structure and its nnorpho-syntactic realization. antworten
——
beantworten
d (E)
auf/daß —- —
—
a —— O
(O)
d —rrrr(E)
answer
A
O b (E)
Ob, that (O)
In this 'two-level1 specification, brackets ( ) , with the lexico-semantic specifications, automatically imply facultative on the morpho-syntactic level.
Agent (A) need not be specified on the morpho-syntactic side, since
it is either in the subject position, and then it is always in the nominative case; or it is in the passive, and then it is preceded by von (or durch). Prepositions unmarked as to case imply the accusative case.
Cases are
specified as they occur, e.g. a (accusative case), d (dative case), and g (genitive case).
The symbol '/' indicates alternative possibilities for
morpho-syntactic realization.
With the above specification, the following
errors could be accounted for: *Bitte, antworten sie mich.
(4.1)
*Bitte, beantworten Sie mir.
(4.1)
*Bitte, antworten Sie auf wir.
(4.1)
*Haben Sie seinem Brief geantwortet?
(4.35)
schon
The possible ambiguity of the sentence (85) Ich beantworte ihm den
Brief.
can easily be resolved by applying relational constants as specified in 2.2.4.5: (85) Ich beantworte* ihm den A
E
Brief. O
(86) Ich beantworte ihm (für ihn) den A
FIN
"I answer the letter for h i m . 1
Brief. O
71
More and more complex semantic sentence plans could be introduced, such as: (87)
Wir antworten ihm mit einem Brief Α
Ε
auf seine: Anfrage.
Ι
Ο
'We answer his inquiry with a letter."
Depending on the communicative situation various structural compressions can occur, e.g.: (88)
Hat er schon geantwortet?
(89)
Hat er dir schon geantwortet?
(90)
Hat er auf den Brief
(91)
Hat er den Brief
for schon geantwortet?
schon beantwortet?
From a communicative as well as a didactic point of view, it is acceptable if the student first uses the sentence that spells out the full information. It is in this context that the argument against G tze (see 2.2.3) gains significance.
Not acceptable, however, are such renderings as:
(92)
*Er hat auf meinen Brief
(93)
*Er hat meinen Brief
beantwortet.
geantwortet.
Let us now turn to the other three verbs: fragen
'to ask 1
befragen
'to question', 'interrogate 1
eine Frage stellen
'to put a question'
From the test items: fragen
(1.9)
A: B:
befragen
fragt Ernst? Polizisten.
(4.7)
They were questioned by the poJice. eine Frage stellen
W
'Sie wurden von der Polizei befragt."
(4.43)
She asked me a question.
'Sie stellte mir eine Frage. 1
72 The following errors were observed: (1.9)
wer/wem instead of wen (23 students)
(4.7)
gefragt instead of befragt (39)
(4.43)
fragte eine Frage instead of stellte eine Frage (16).
HuS paraphrase fragen with eine Frage stellen and enter fragen and befragen äs: f ragen
Sn
=
bef ragen2+(1)=3
(St:)
Sn Sa
Student errors indicate that the riajor clif ficiilty w.is the fragen/befragen distinction (39) , followed by assigning the correct surface case (23) and translation of ask a question as eine F^age fragen (16) . The first two problems are resolved by the same specification as applied to antworten/ beantworten. a
fra en
*
befragen
—
nach/ob, w
)
r
--
eine Frage stellen
A
( >)
Surface case specification with certain prepositions is not necessary since they are always followed by that case, e.g. nach in fragen above is automatically followed by dative case; otherwise the required case is indicated, e.g. (a) . The specifications above capture all errors made by those tested. As far as eine Frage stellen is concerned, this has to be learned as one unit. We will later (3.2.9) have an opportunity to reflect on the semantic status of such verbs as stellen in a discussion on "function verb constructs ' . Comparing the two verbs antworten and fragen it can be observed that the E with antworten is realized by the dative; and with fragen, by the accusative case. No such distinction is evident in English: (94)
He asked me.
(95)
He answered we.
73
If befragen is used with the reflexive, it takes on a new meaning.
Hus do
not have a separate entry but mention this possibility under bc.fmtjcn
(1973:
1O5) as one of the prepositional choices, which is confusing. sich befragen ht?i
Λ
Ε
(Ο)
inquire
A
O
(t:)
(erkundigen)
It is not adequate to consider this reflexive the object of the nonreflexive befragen.
If this were the case, a new relational constant for the "bci-
phrase" would be needed.
This "bet-phrase" is, however, the morpho-syntactic
realization of the same relational constant as the surface object with the nonreflexive befragen. (96)
I questioned him about the 'Ich befragte ihn
(97)
I inquired
theft.
ber den D i e b s t a h l . '
(with him) about the accident.
'Ich befragte Unfall.'
(erkundigte) mich bei ihm
ber den
Contrasting the two verbal concepts underlying sieh bafragr.n bei and -inquire, we see that they are instances of different semantic sentence structures: . , , , sich befragen , ι w , (erkundigen)
inquire (98)
— A
, . bei
—— A
after/about Ο
Ich befragte
bei
nach/
E
ber(a)
(0)
of (Ε)
(erkundigte)
mich bei
ihm nach dem Weg.
Ί inquire the route (of h i m ) . ' (99)
Ich erkundigte mich nach der Adresse.meines 'I
Freundes.
inquired about my f r i e n d ' s address.'
With sich befragen bei, the E is obligatory; inquire seems to focus on Ο (what the inquiry is about).
Inquire above is, of course, only to be inter-
preted in the sense of find out about something (or someone).
The more
frequently used German verb for inquire in the above sense is sich erkundigen. This verb requires the same relational constants as inquire. The English inquire into, which is a separate verb from the one just discussed, has the
74
German equivalent (equivalency is to be understood in a wider sense): erfragen
a/ob,w
~7~
bei von
Ο
(Ε)
A conparison of the semantic sentence structure of such verbs as fragen, befragen, erfragen reveals interesting insights into natters of focus; fragen 'ask1 appears to be neutral; befragen
'question1 has the focus on E;
'inquire1 has the focus on O.
erfragen
frage"1
A
(E)
(0)
ask
A
(E)
(0)
befragen
A
E
(0)
question
A
E
(0)
erfragen
A
0
(E)
inquire
A
Ο
(E)
Another observation that can be made from such a comparison is the multiple function of the accusative in German; with fragen and befragen, it is the morpho-syntactic realization of E; with erfragen, it is O. This investigation into verbs related to fragen or ask could be continued
(anfragen,
ausfragen, in Frage stellen, etc.) in order to gain insight into the different logico-semantic valences of these verbs.
Our present task is only
to indicate the possibilities ana limitations of a simultaneous consideration of logico-semantic valence and its morpho-syntactic realizations (syntactic valence). 'Verbs of saying1, including verbs like ask, say, speak, tell are considered action-process expexiencer verbs by Cook (1972:23).
The verbs
answer and question are entered as derived Ά Ε Ο verbs' with the case frame A E (Cook 1973:67, 73).
r
0 lex +| 1
ι /
The 0 is lexicalized in question:
+[
(~~~
A,E\ add
J
A}A30*\/0 lex. It is not clear what really is meant by 0 since
... all verbs of saying imply a speaker (A), a hearer (Ε), and what is said
(O)' (Cook 1973:73). In our context, O is not considered lexicalized, but stands for what is being asked (said), as in: (1OO) I questioned him (about the
Α
Ε
Ο
accident).
75
The observations we made with this first small group of verbs concerning the relational constants Α Ο Ε have shown the availability of the concept of logico-semantic valence as a canton basis for coitparing related verbs within one language and across languages.
The formal presentation of val-
ence specifications on two levels is explicit enough to supply necessary information for the morpho-syntactic realization of basic semantic sentence nuclei.
Especially in cases such as answer and antworten/beantworten, where
the meaning range of one verb in English is shared by two verbs in German, our method of specification states the important differential information in a simple way, that is semantically relevant though not exhaustive. 3.2.1.2
Sprechen/sagen/'reden/erz hlen; "to speak'/'to say1/ 'to talk'/'to tell, narrate1
This is an interesting group, especially when compared with their English equivalents.
Even though only morphological mistakes were made in 1.2O
(Part 1 of test; Item 2O: sprechen), an investigation into the valence characteristics of these verbs is warranted. The entry for sprechen in HuS (4O5) is: sprechen
1+(έ) "J
Sn
(Sa/pS)
This is wrong, it should read: Sn (Sa/pS)
(pS)
The following possibilities for pS are stated: p = mit, p =
zu, vor
ber, von
p = f r, gegen p = aus
The various possibilities available with the prepositions above are of major concern here, since the grouping of prepositions reflects their different functions in regard to relational constants. This is vaguely indicated in HuS's footnotes (4O6). Incidentially, we cannot agree with HuS when they state that pS (human) as in mit dem Freund can be substituted by Sa (human) as in den Freund.
There is a fine yet
significant semantic-pragmatic difference between these two renderings. (1O1)
Ich spreche mit dem Lehrer. ' I speak with the teacher.'
76 (1O2)
Ich sprec/ie zu dem Lehrer. speak to the teacher. 1
(
'.))
Ich sprticht: vor den Lauten. 'I speak before the people. 1
The prepositional phrases with mit, xu, and vor are instances of E. are comparable to English with, to, and before.
They
The prepositional phrases
introduced by über and von in German as well as those introduced by about or of in English reflect 0: (1U4)
Kr sprach über den Lehrer.
A ü
'He apoke about the teacher.' (105)
Sie sprach von der Reise.
AO
'She spoke about ( o f ) the t r i p . ' (106)
Sie sprachen von dem Unfall.
A 0
'They spoke about the accident.'
A combination of one fron each group is possible: (107) Er sprach mit dem Lehrer über den Unfall.
A EO
'He spoke with the teacher about the accident.'
but not two of the sane group: (108) *Er sprach über den Lehrer von dem Unfall. *'He spoke about the teacher of the accident.'
unless we conjoin them by und: (109) Er sprach über den Lehrer und von dew
Unfall.
"He spoke about the teacher and the accident. 1
Considering the third group of prepositions entered by HuS, we found that they not only introduce phrases referring to O, but also include such notions as the position taken towards O, or judgments.
No subclassifica-
tion of the 0 as attempted in the theoretical section can account for this relational constant.
It could be argued that this sprechen is actually a
different verbal concept, something of the sort be for (against). to the prepositional phrase with aus Out (HO)
Er spricht aus
of:
Erfahrung.
'He speaks out of experience."
And as
77
we have to ask whether this example can be paralleled by: (111)
Er spricht laut. 'He speaks loud.'
(112)
Er spricht lange. 'He speaks long."
(113)
Er spricht überzeugend. 'He speaks convincingly.'
Can it be considered a free adjunct? Aus Erfahrung above is to be considered such a modal specification. It is interesting to note that with aus, für, gegen no A is necessary in the semantic sentence structure, which is further indication that we are dealing with another verbal concept. (114)
Wahrheit sprach aus seinen Morten. 'There was truth in his words.'
(115)
Die Fakten sprechen gegen ihn. 'The facts speak against him. 1
This observation shall not be further pursued here.
Sprechen then can be
stated as:
sprechen
—r—
-
mit,zu,vor l über(a), von/daß ··' (E) l (O)
and compared to
speak
_
__ / _ with,to,before I _ about,of
So far, the two verbs sprechen and speak seem not to pose any particular problems.
The symbol '
tants must occur.
/
' indicates that one of the facultative ac-
This specification does not correspond to HuS's entry,
which is based on syntactic valence only. A certain comrnunicative context can make the omission of E and possible, but not without the participant's
78
knowing something about the particular situation.
Nevertheless, sprechen
seems to be somewhat neutral in regard to E and O, maybe Oriented towards the speaking act itself
(Anderson 1971b:146).
A verb that is frequently
exchangeable with sprechen is reden. HuS
reden
Sn (Sa/pS/NS
daß)
pS is specified
as p = mit,zu,vor p = von, über
Ihe difference in facultative valence between sprechen and reden as stated by HuS and the "MS daß" possibility with reden are not justified when considering examples like: (116)
He spoke with him about the accident. 'Er sprach mit ihm über den U n f a l l . 1 'Er redete mit ihm über den U n f a l l . 1
(117)
He talked about the fact
that she will be coming soon.
'Er redete davon, daß sie bald kommt. 1 (118)
He spoke about the fact
that she will be coining soon.
'Er sprach davon, daß sie bald kommt.'
The English talk corresponds very closely to reden, especially when one considers the minute stylistic differences between sprechen and reden, i.e. a slightly different degree of formality: He gave a speech to the association. He gave a talk to the association. (Anderson
1971b:145)
In the English verb forms speak/talk this difference appears to be neutralized. HuS
sagen
/,._,
Sn
Sa/NSdaß,ob,w,
(Sd/pS)
0 with sogen is typically of the sentential type, i.e. in itself a sentencesemantic content: (119)
I said that I will come. 'Ich sagte, daß ich kommen werde.'
79
Sagen is 'content'-oriented with obligatory 0.
The E is realized either by
dative or prepositional phrase with zu. sagen
_
a(über,von)/NSdaß,w ,_L
Ob(about)/Cl
that,w
d/zu _
to ( )
This specification accounts for such sentences as: (12O) He said something about the accident to me. (He told me something about the accident.) 'Er sagte mir etwas über den U n f a l l . " (121)
He said something. 'Er sagte etwas. 1
In the latter exanple etwas is actually a proform. (122) Er sagte zu mir, daß ... 'He said to me that . . . '
Notice that the ccnbination of the "superficial double object1 (something about) cannot be contained with E realized as a prepositional phrase with zu: (123) *Er sagte zu mir etwas von dem Unfall. 'He said something about the accident to m e . '
Sagen does not permit adverbials of the type: (124) *Er sagte lange. *'He said long. '
but (125) Er sprach (redete) lange. 'He spoke (talked) long. 1
is possible. oriented.
This is another indication that sagen is 'content* (O)-
80
The verb tell was already used above as a synonym of say in: (126)
He told me something about the accident.
While in German the omission of E is possible with sagen, this is not the case with tell: (127) Er sagte etwas
ber den
Unfall.
*'He told something about the accident. 1
This is the reason why sagen was related to say, since say also permits omission of E, ignoring, of course, the different surface realization of E if present. Tell is definitely Ε-oriented, therefore: tell Ί
Ob Ε
Α
Ob(about)/Cl (Ο)
that,w
accounting for sentences like: (128) He told me. (129) He told me the truth. (130) He told me something about ... (131) He told me that ...
but not (132)
*He told.
(133)
*He told something about ...
The German sagen can be rendered in English either by say or tell, but only within their respective valence requirements: sagen
A
O
^.say (E)^^
A
o
(E)
Α
Ε
(Ο)
^
(134) Er sagte etwas A
ber den
Unfall.
O
'He said something about the accident. 1 *'He told something about the accident.'
81
The (*) example is excluded due to the above valence specification of tell. Such verbs as sagen, sprechen, and reden are prime examples of the intricate patterns that exist between logico-semantic relational constants and morphosyntactic realizations; the very simple specification above makes clear the significant difference between sagen and say/tell, which is a source of frequent errors. How would this logico-semantic valence account for examples with asterisks as reported in Zarechnak (1974)?
Consider:
(134a) *I spoke something about that to John. Speak does not permit 0 to be realized as 'Ob+about1 and: (135)
*John said.
is in violation of say A 0 ( E ) , since 0 is missing. (136)
So also:
*John told.
violates tell Α Ε (Ο) .
It should be mentioned that (135) and (136) .could occur
in a certain context but only with contrastive stress. (137)
*I spoke to John that I would come.
The realization of 0 with a 'Cl that' is not possible with speak; it would be possible with tell, resulting in the correct sentence: (138)
I told John that I would come.
(139)
*He said for
But in:
two hours.
the obligatory 0 for say is missing.
The focus is on the Ο (content), and
any modal imposing a Time element (T
J upon the verbal concept results in
awkwardness. (140)
Such as: *He said something about the accident for two hours.
This is not the case if Time ('Ik,nct) is added to the entire sentence-semantic content as in: (141)
Yesterday, he said something about the accident.
82
It is interesting to note that with speak a tine concept ( +^ ) agrees perfectly, due to the fact that speak is neutral in regard to or E, and is "activity'-oriented: (142) He spoke (to us) (about the accident) for two hours. Tell is specified in Cook (1973) as follows: tell
+\
A,E,(0)\
change (O) to 0 + \
A,E,O\j O deletable
Ihis tell can only be understood as specified above. There is, however, another tell (narrate). In: (143) Mother told the child a fairy
tale.
'Die Mutter erzählte dem Kind ein Märchen.'
is not deletable.
This tell2 has the following specification:
tell (narrate, report)
——
ob
Ob/to
As can be seen from the specification above, this tell- corresponds to the German erzählen. It is not -oriented like tell^. HuS
erzählen
erzählen (berichten)
Sn (Sd) (Sa/pS/NS
f2)=3
daß,ob, w)
a,über,von A
( )
The point to be made here is that any conflation of "case frames' , as neat as it may look, conceals semantic as well as morpho-syntactic differences. Thus, for example, is d with erzählen, and tell- does not need E, as does
(144) He told a story. 'Er erzählte eine Geschichte. ' (145) *He told about the accident. 'Er sagte etwas über den U n f a l l . 1 Erzählen is O-oriented, so is tell,,·
83
To conclude the discussion on this group of 'verbs of saying1, let us consider the following sentences: (146)
Er sagte etwas über seine Reise nach Europa. 'He said something about his trip to Europe. 1
(147)
Er sprach über seine Reise nach Europa. "He spoke about his trip to Europe."
(148)
Er sagte uns etwas über seine Reise. "He told us something about his trip. 1
(149)
Er erzählte von seiner Reise. 'He talked about his trip. 1
(150)
Er berichtete über seine Reise. 'He reported on his trip. 1
Zarechnak (1974), in his interpretation of say, speak, tell, and talk makes the observation that: ... these elements show a gradual movement from articulate words, no matter how logically disconnected, to connected discourse of varying length. (Zarechnak 1974:36)
This, to a certain degree, is reflected in the logico-semantic valence specification of these verbs. The task of this rather detailed investigation was to capture some of the differences and similarities between these verbs as they cause difficulties for the learner of a foreign language, in our case, German and English. Considerations on different levels are necessary if one wants to arrive at a comprehensive understanding of valence. In sunmary, the following list reiterates the logico-semantic valence configurations of the verbs discussed above: sprechen
A (E / O)
speak
A (E / O)
reden
A (E / O)
talk
A (E / O)
sagen
A O (E)
say
A O (E)
erzählen
A O (E)
tell 1
A E (O)
tell
(narrate)
A O (E)
84
3 . 2 . 1 . 3 grüßen/begrüßen/grüßen lassen/empfangen 'to greet '/'to give regards to '/'to welcome1 Test Item 1.31 dealt with begrüßen. ••a grüßen
—
-^
greet
—
begrüßen
——
——
greet
(as above)
A
£t
The student had to decide on the prefix. the .fee-prefix. Test Item 4.1O read:
—
Fifteen students made errors with
Give my regards to your parents.
Interference from the mother tongue could be observed among responses to Item 4. 1O, e.g.: *Erinnern Sie mich zu Ihren Eltern. 'Remember me to your parents. 1 *Gib meine Gedanken deinen Eltern. 'Give my thoughts to your parents. 1 *Gib meine Grüße deinen Eltern. 'Give my greetings to your parents.'
To your parents was rendered as: *zu Ihren Eltern * Ihren Eltern *auf Ihre Eltern *nach Ihren Eltern
Admittedly, there were different ways of expressing this sentence in German; the following were acceptable: Schönen Gruß an Ihre Eltern! Bitte grüßen Sie Ihre Eltern von mir. Grüßen Sie mir bitte Ihre Eltern.
85
Begrüßen appeared as: *Begrüßen Sie sich deine Eltern. *Gib Ihren Eltern meine Begrüßung. *Begrüß mich zu deinen Eltern.
There were more than ten other iitpossible interpretations of the sentence given in 4.1O.
Mo speculations are made here to account for all the
different interference problems. Of concern is the uncertainty in the use of grüßen/begrüßen. It must be emphasized that the student had no clue to the choice between these two verbs. What is wrong with using begrüßen? As indicated above, it involves the actual communication between the one 'greeting' and the one 'greeted1 in the specific sense greet—welcome. In 4.1O, a rather complex sentence-semantic content is expressed including C and FIN. C refers to the one requesting the message, in this case indicated by the imperative. as well be expressed with lassen'. (151)
Lassen Sie mir Ihre Eltern grüßen.
FIN is expressed in
(152)
Grüßen Sie wir (für mich) Ihre Eltern.
The sentence semantic content underlying: (151)
Lassen Sie mir Ihre Eltern grüßen.
could be illustrated as follows: lassen someone greeting someone ""FIN
for someone
It could
86
There is a very close relationship between begrüßen and empfangen It seems best to treat this empfangen empfangen
(1.44).
"receive, welcome" separate from
'receive something":
HuS
V
empfangen
empfangen
(als Gast)
a
——
1
empfangen „
. , , welcome, receive
A
ti
——
——
A
Sn Sa
'receive, take'
Cx
A
—— n
Ob —-— £j
——— L·/
This specification also allows us to show the difference between: empfangen
A 0
'receive'
('take',
'receive'
('get')
'accept')
and bekommen (erhalten)
DIR
O (DIR )
The little help that dictionaries provide in such cases is illustrated by: erhalten
= bekorrmen,
empfangen
bekommen
= erhalten,
empfangen
empfangen
= erhalten, bekommen (Mattutat 1969)
3.2.2
On the Benefactive notion
3.2.2.1
State/Benefactive
The errors in 1.11 and 4.3O concern the surface case of B as well as that of 0; translation of to in to whom resulted in prepositions like an, zu, bei, von. HuS
gehören
gehören
(153)
—
(als Eigentum haben)
—s
'belong'
Das Buch gehört mir.
'The book belongs to me. 1
Sn Sd
—
-j~s
87
This siirple specification accounts for the mistakes made in the test. Any attenpt to ignore the translation factor with adult learners of a foreign language would be shortsighted.
It is suggested that specifications like
the one above can help to eliminate such mistakes if introduced from the very beginning and used in a repetitive and consistent way throughout a foreign language program. What would own and the German besitzen look like?
They express the same
sentence semantic content but with a different focus: u besxtzen
— s
a
ob
—
own
_____ „s
The relational constants have remained the same with the exception that the subject selection has been reversed; this is clearly shewn in the specifications above. Gehören can be contrasted with angehören, which does not oxpross possession but the concept be a member of, e.g. HuS
angehörten
Sn Sd
(154)
Er gehört keiner Partei jn. 'He does not belong to a p a r t y . '
(155)
Der Hund gehört der Klasse drr Säugetiere an. 'The dog is a mammal.'
... angehören y
d /'art
belong . . mrmbpr of')
...... Part
to
A frequent error in translating from English to German is: *Er gehört der Partei. even when the student knows that belong to should be -rendered by the dative. The participation of the prefix an- in the total verb meaning must be indicated.
HuS distinguish no less than five variants with gehören (angehö-
ren is treated as a separate verb), and suggest a certain hierarchy in difference so that differences on level two and one signal more prominent 'meaning differences' than differences on level three (HuS 1973:155).
This
observation is correct, but the variant idea can be misleading, since the difference in meaning is so prominent that it is preferable not to speak of variants at all, e.g.: sich gehören
'to be proper, becoming 1
gehören (gebühren)
'to deserve"
gehören (nötig sein)
'to be necessary'
The difficulties the learner encounters with these homonyms can be handled more properly by using individual valence descriptions for each variant. These specifications (logico-semantic and morpho-syntactic) are considered supportive for the understanding of similarities and differences between such variants and their counterparts in the learner's native language. One of HuS's variants will be included here since it corresponds to belong to in appearance. HuS
gehören
(156)
Sn pS
Arlington gehört nicht zu Washington. "Arlington is not a part of (does not belong to) Washington." zu
gehoren
--
Part
belong
to _- __ Part
^
In English the distinction "A 0" in belong to (re part of, be a member of) is neutralized on the surface. (157)
He belongs to the soccer team
'angehören 1
(158)
Georgetown belongs to Washington.
'gehören zu'
There is,
finally, a belong to that reflects the
, .. gehören
( 1 .59 )
——
o
in, auf ... -
L.
Dieser Teppich
'belong' , .
(go)
gehört: jn.s Wohnzimmer.
"This rug belongs in the living room.'
——-
o
in, on --
89
In surmery: besitzen
B
geh ren
O
B
geh ren zu
O
O
s
geh ren
υ
3.2.2.2
Ο
belong to (be part of)
Ο
Ο Part
belong (have its proper place)
Ο
L
belong to (be a member of)
Ο
Ο
belong to (possession) Part
L
Dir
angeh ren
Β
Pur t
Dir
Part
Process/Benefactive
In Test Item 1.25, with bekommen, the prefix was not given.
Even though the
communicative context indicated the meaning of the verb, a whole group of prefixes appeared in the answers, e.g. on-, vor-, aus-. HuS: bekommen. 2+(l)=3
Sn Sa (Part
2/Inf)
(erhalten)
bekommen (erhalten)
DIR„ G
a
von
0
(DIR) S
Ob
(160)
from
Sie bekam einen Brief
von Peter.
'She received a letter from Peter. 1 (161)
Hast du das Geschenk schon bekommen? 'Have you received the present already? 1
HuS1s description of bekommen would include: (162)
Man bekommt hier
pfel.
One can get (buy, obtain) apples here. 1
90
which is considered a separate verb in this investigation. bekommen (kaufen) (163)
~Ä~
a ~Ö"
auf, in . .. Z
Briefmarken bekommen Sie nur auf dem Postamt. 'Stamps you can only get at the post office. 1
(164)
Kann ich noch ein Zimmer bekommen? 'Is there still a room available?' still get a room?')
('Can I
Even though L is omitted in this last exanple, it is conceptually present. Let us now consider HuS's exanple: (165)
Er bekommt Angst
(Hunger).
'He is getting afraid (hungry). 1
There is no Benefactive relation involved here but rather an Experiencer. In Test Item 4.2 the students were to render in German: J am getting hungry. Twenty-nine students failed to give one of the two possible German sentences: (166)
Ich bekomme Hunger.
(166a)
Ich werde hungrig.
The main difficulty was getting, which was stated as sein or hoben, e.g.: *Es ist mir hungrig.
This shows that the student has mistakenly applied the structure of such sentences as: (167)
Es ist mir kalt. am cold.'
(168)
Es wird mir schlecht. am getting sick.
The specification for the verb: bekommen
—
-~-
91
indicates the inchoative process.
We notice different possibilities in
English: (169)
I was getting hungry.
(169 )
I grew hungry.
(169b)
I became hungry.
It is this last example with become that adds to the difficulties the English learner already has, because of the formal similarity between become and bekommen.
This results in transfers like:
(17O)
* Ich bekomme hungrig.
A full investigation of verbs like JTOU, turn, and beuo>-ie is not possible within the scope of this thesis.
The important observation to be
stated here is that bekommen can be the surface form for totally different sentence semantic contents, e.g.: bekommen
DIR
O
bekommen
A
O
bekommen
E
O
I
G
(DIR ) S
L
get
(receive)
g e t (buy) get
(become)
In Test Item 4.25 yet another meaning of bükomncn was called for: He likes to eat fish, but it
does not agree with him.
'Er ißt gern Fisch, aber Fisch bekommt ihm n i c h t . '
The variety of answers given was simply amazing; a few may suffice: * ...
, aber das geht nicht bei ihm.
* ... , aber es tut ihm weh. * ...
, aber es macht er krank.
* ... , aber er geht etwas kränklich. * ...
, aber das wirkt ihm nicht gut.
* ... , aber es tut ihm schlecht. * ... , es paßt ihm nicht.
92
The following specification for this bakotmian is suggested: with (xu.it (food) (171)
the liealth
of)
's
Die IJostinmi I.ah bekommt dem Kind
nic-fit.
1
'The canned m i l k doc. ;; not agree w i t h the c h i l d . 1
but (172)
Das Kind bekommt das üssen. 'Thu child q o t y the m n a l . 1
3.2.2.3
Action/Benefactive
In Test Item 1.7, with dunkln 'to thank", eighteen students made mistake£ with DIR_, most of them because they used the wrong case. HuS
V
danken (Dank
danken
.. , thank
/co
-5ΊΓG
Ob
—
94
In Item 4.9 the following was presented: He was given a new car. No less than 32 students used the DIR,, as subject in the German sentence: *Er wird ein neues Auto gegeben. *Er wird einen neuen Hagen gegeben.
Ihe fact that DIRr cannot be subjectivized in German is not included in the valence specification, but can be explained on semantic grounds as a general principle of German gramar.
This also applies to 'verbs of saying1 (as
discussed above in 3.2.1.2), if they have a facultative E, e.g.: (178)
i'r aac/te mir etwas über den
Unfall.
(17öa)
Mir wurde etwas über den Unfall
(178b)
*Ich wurde etwas über den Unfall
Passive gesagt. gesagt.
'I was told something about'the accident. 1
Ihe following diagram shows this difference between English and German: give
A
DIR
geben
A
active
active
passive 1
passive 1
passive 2
(179)
John gives the girl the book. 'John gibt dem Mädchen das Buch.'
(179a)
The book is given to the girl.
"Das Buch wird dem Mädchen gegeben.1 (179b)
The girl is given the book.
DIR
95
As expected, the same difficulties were encountered with schenken give a present', in 1.18 and 2.2, where 43 students made a mistake in the morpho-syntactic realization of DIRr: HuS
schenken
,.,_
Sn Sa
d
schenken
R
jR
G
(Sd)
a —
give a present
A
Ob/to DIR , 6
Ob ~ ~
This specification permits a more homogeneous subclassification of verbal concepts; geben, schenken, zeigen, etc., could all be entered with the sane sentence- semantic content and are, therefore, subject to the same subject selection rule. Ihe bring en/nehnen 'bring* /'take' field:
The example in Test Item 4.4,
with bringen was: Was he not taken home? 'Wurde er nicht nach Hause gebracht?'
The following verbs along with bringen were used in the answers: kommen, begleiten, abholen, holen.
The Lp.
appeared as zu Hause instead of
nach Hause. HuS
V
bringen
, . bringent
—
Sn Sa pS
a —
in,auf
...
Dir
^ , take, Ί (ISO)
Α
Ob Ο
to,into ... L .
Dir
The ambulance took him to the hospital. 'Die Ambulanz brachte ihn ins Krankenhaus. 1
(181)
*The ambulance took him. *'Die Ambulanz nahm ihn.'
The obligatory pS corresponds to L^. .
nehnen,
96
A second verbal concept bringen.·, can be distinguished from bringen... HuS
V
bringen
, . bringen. 2 (182)
-— Ά
Sn Sa
, ·,·,--, a — Ο
(sd^
d , (DIRJ
, . bringy
/nrn
— Λ
G
Ου 0
Ob/to (DIR ) G
Er bring mir das Buch. 'He brings me the book.'
In a sentence like: (183)
Er brachte mir den Brief
zur Post.
'He took the letter to the post o f f i c e for m e . '
the dative can, of course, only be understood as a realization of FIN. On the other hand, in: (184)
Er brachte mir den Brief
ins
Schlafzimmer.
'He brought the letter to me while I was in the bedroom.'
mir is an instance of DIR„. (j To account for so many errors with bringen/'take','bring', the following specification is suggested: bringen, 3 l
A
O
L . Dir
take
A
O
L„. Dir
bringen^
A
O
(DIR}
bring
A
O
(DIR )
2
G
G
it is understood that this specification of bringen does in no way cover all the "meanings' this verb form can have. A further complication is caused by the verb nehnen. less than ten variants with nehnen. M text.
and V,- are of interest in this con-
With regard to: HuS
V
nehmen
Sn Sa pS
this nehmen. corresponds to bring en„ above. Re: (185)
HuS's
V
nehmen
HuS mention no
Sn Sa (Sd)
Er nimmt das Buch vom Regal. 'He takes the book from the shelf. 1
97
(186)
Er nimmt das Buch nach Hause. "He tcikes the book home. 1
this corresponds to: nehmen . 2
take
a O
— A
2
von,aus/nach ...
Ob O
A
from/to
...
The following sentence semantic contents may be compared: (187)
Er nimmt dtun Kind das Buch vom Regal.
AOL
(SIN)
"He takes the book from the shelf for the c h i l d . ' (188)
Er nimmt das Buch nach Hause.
AOL.
'He takes the book home." (189)
Er nimmt dem Kind das Buch (wt?g) .
A O DIR J
'He takes away the book from the child.' The following types of nehmen can be distinguished: nahmen, l
A
O
L
. Dir
take
nehmen2
A
O
(L/LD · )
wegnehmen
A
0
(DIR )
A
O
take
A
O
(L/L
take away
A
O
(DIR )
l
L
. Dir
Ο
. )
Ο
The distinction between ΟΙϊ^, . and 1^.
is often not easy, but within the
present framework of logico-semantic valence it is still a useful one. In Test Item 4.13: From whom did he buy that old car? 'Von wem hat er das alte Auto g e k a u f t ? '
sixteen students had problems with DIR-J and the selection of the verb kaufen; verkaufen and einkaufen were used instead. kaufen
—
a — O
von /nrp
,
(DIR )
buy3
A
Ob
from
O
(DIR )
98
becomes obligatory in the German verb abkaufen. (190)
Kr kaufte mir den Magen ab. 'He bought the car from me. 1
The prefix ab- serves as a focusing device on DIR„ but does not by itself include the entire DIR„ concept. HuS (191)
verkaufen
,M^,
Er verkaufte
Sn Sa
(Sd/pS)
einen Wagen,
'tie sold a car.' (192)
Er verkaufte einen Hagen an Herrn Bauer. 'He sold a car to Mr. Bauer. 1
(193)
Er verkaufte Herrn Bauer einen Hagen. 'He sold a car to Mr. Bauer.'
A O DIR
G
or 'He sold a car for Mr. Bauer.'
, ,. verkaufe,^
—
a —
d/an (
A O FIN
,, sell 1
—
Ob —
Ob/to
C
Tb illustrate the limits of the present framework of specification the following examples are of interest: (194)
Do you sell
cosmetics?
'Verkaufen (führen) Sie Kosmetikartikel?' verkaufen_ (führen) (195)
A
O
sell (carry)
He do not carry this item. 'Wir führen diesen Artikel nicht.'
And even: (196)
This toothpaste sells well. *'Diese Zahnpasta verkauft gut. 1
)
G
99
In German the realization of 0 as subject of an Action-Process verbal concept is problematic. The valence of verkaufen cannot be reduced to one place. Even in English some irodification is necessary in these cases (at least conceptually): (197)
(*)This toothpaste sells.
In Gentian, the non-realization of the A results in reflexivization. Diese Zahnpasta verkauft
sich gut.
Rrinkmann (1971:208) considers this sich a 'passive reflexive1. Without entering the controversial discussion on degrees of reflexivity, it can be observed that the sich in the example above completes what otherwise would be a deficient surface realization of a two-place predication. It could be argued that in cases like the following, the modal specification is part of the semantic valence, i.e. obligatory. (196)
This toothpaste sells well.
(198)
This material washes easily. 'Dieser Stoff w scht sich leicht. 1
(199)
This book reads easily. 'Dieses Buch liest sich leicht. 1 (Zimmermann 1972b:177)
All the examples above can be paraphrased in German by using Lassen or sein + infinitive with zu. (2OOa)
Dieser Stuff
l
f2OOh)
Dieser Stoff
ist
t sich leicht waschen. leic-lit zu wasclien.
There are instances where the reflexive is not sufficient to render the O in English; that is the case when the subjectivized Ο is quantified by another surface object, as in: (201)
The novel sold ΙΟΟ,ΟΟΟ
copies.
Ihe only possible solution in German is a passive sentence: (202)
Von dem Roman wurden 1OO OOO Exemplars
verkauft.
100
Returning to the verb kaufen we observe: (203)
Er kaufte
seiner Tochter einen Porsche.
'He bought his daughter a Porsche.' (204)
Er kauft
sich einen Porsche.
'He buys himself a Porsche. 1
In these, when DIR is coreferential with A, the DIR„ is realized as re(j u flexive in the dative case. This could be stated as: kaufen
d/fur - ' „ .
—-
(DIR
—-
)
DIR,, G
Refl
d
DIR
L.....J and is to be read as:
If DIR
= A, then DIR_ is realized as dative reV.3
(j
flexive. In German there are other verbs for to buy for· oneself, such as: anschaffen
A
DIR,,
O
(The uninterrupted arrow indicates the obligatory reflexive; the interrupted arrow, facultative reflexive.)
In accordance with this specification, there
is no actual sentence possible other than: (2O5)
Er schaffte
sich einen Porsche an.
The coreferentiality between DIR anschaffen
is obligatory.
and A in the semantic sentence content of
The same holds true for zulegen which is a more
colloquial rendering of anschaffen.
The specification above seems to
capture more adequately the constraints on the composition of sentencesemantic contents than König's (1972) observation: Diese Tatsache erklären wir am besten als Folge einer Restriktion, die festlegt, daß Verben wie zulegen und anschaffen obligatorisch Identität zwischen Subjekt und indirektem Objekt verlangen. (König 1972:37)
101
A very similar verb beschaffen (2O6)
is of interest here:
Ich beschaffte
meinem Freund die Bücher.
' I got the books for my f r i e n d . ' (2O6a)
ich beschaffte
mir die Bücher.
got the books for m y s e l f . ' beschaffen
A
DIR
L____J While anschaffen imply pay for.
G
O
implies obtain and pay for, beschaffen
does not necessarily
In summary, we have the following simple and yet revealing specifications: kaufen
(DIR )
(DIR )
O
buy
U....J abkaufen verkaufen
buy from
DIR
sell
(DIR
G}
anschaffen
DIR,
zulegen
DIR
beschaffen
DIR
O
get for oneself
buy for
O
get
(buy)
oneself
(obtain)
It is understood that this model just presented could be applied to numerous small subgroups of verbs and contribute to a better understanding of logicosemantic constraints imposed by certain sentence-semantic contents. This kind of information is not found in dictionaries. Although the following verbs were not included in the test, they are discussed here because they illustrate the usefulness of the logico-semantic
102
valence concept. Let us consider first the 'rob1/'steal' pair. According to Cook (1973), 'rob'/'steal' are examples of Action-Process/Benefactive verbs: steal
+\
A,(B),O\
change (B) to B
+\
rob
+\ I
A,B,(O)\ |
change (O)
+\ A,B,O\/O ( —— J
to Ο
Λ,Β,θΙ/
deletable deletable
(Cook 1973:74) Of special interest for this investigation is the following observation: All of the verbs of transfer imply a source, a goal, arid an object transferred. If the Α-case represents the goal, as in the verbs: rub, steal, tho B-case represents the source. Some verbs, like steal, demand that the A- and O-case be expressed but leave B-case optional in surface structure. Some verbs, like rob, demand that the A- and B-cases be expressed but leave the O-case as optional in s u r f a c e structure. (Cook 1973 :7 r i)
Comparing the above-stated verbs with the corresponding German verbs rauben/ stehlen, we observe that different lexicalization processes take place: He stole thv car (from
his neighbor).
A
(DIR )
'Er stahl das Auto
(von seinum Nachbarn)
A
O
He robbed
(of
A
(O)
A
O
J
(2O8)
bank
the- money;.
'Er raubte das Geld (von der
Bank).'
(DIR ) S
DIR S
(DIR ) O
While for the English rob, O is facultative; for the German rauben, 0 is obligatory and (DIR_) is facultative. This changes as scon as the prefix be- is added to rauben/stehlen. Now both require DIR^ but have O as a facultative actant. The focus in rob is on DIR^; in rauben, on 0. The heprefix brings about focus on DIR,^ and includes the concept of mething being r.tolr.n = 0. It is suggested that ba- carries the meaning which 0 has in.the semantic sentence content: A DIK,,: > ι
ra ben
Λ
berauben
Λ
(DTK
ο rob
'-hier
A
>.·; tehl^n
Λ
steal
A
103
Such a specification reveals a number of areas for potential errors which could occur in the obligatory/facultative actant assignment or in the morphosyntactic realizations.
In the above specifications the morpho-syntactic
realizations have been omitted because they are basically the same as with kaufen.
3.2.3
Two examples for possible errors may suffice: (209)
*Er rauhte die Bank.
(210)
*5ie hat dem Vater bestohlen.
Application of logico-semantic valence to locative verb types
The following verbs from the test will form the basis for discussion: abfahren
i.23
'depart 1
abfliegen
1.29
'depart (by p l a n e ) '
fliegen
4.6
'fly'
abholen
2.6
'pick up 1
aussteigen
1.27
'get off (bus, train, . . . ) '
besteigen
4.40
'climb'
gehen
1.2, 2.3
'go, drive'
hängen
1.4
'hang'
legen/liegen
1.35
'lay/Lie1
Platz nehmen
4.18
'take a seat 1
setzen/sitzen
1.37, 1.49
'seat/sit'
stehen/stellen
1.5O
'be located/place, put 1
treffen
4.18
•meet 1
wohnen
1.6
'live (reside)'
Criteria for the distinction 'obligatory/free' in connection with prepositional phrases were discussed earlier (2.1.3.2).
104
3.2.3.1 On the State/locative relation With stehen, HuS specify: V
stehen
Sn (pS)
=
This V.. is only possible when Sn is +Human. L is obligatory, as in Test Item 1.5O: A •.
B:
If 0 is not specified as +Human,
· ·· Ja, da stehen auch die ander en_ Räder. "Yes, there the other bikes are (standing). 1
in, auf
stehen
.. .
s
The Gernan verb stehen in the sense to be located implies upright position for objects entered in 0 . S
(211)
Die Vase steht auf dem Tisch. 'The vase is
(standing) on the table,'
The OS L relationship in German requires the dative case on the surface. HuS enter liegen as: HuS
V}
liegen}+(1)=2
Sn (pS)
V2
liegen2
Sn pS
Only the V- is a locative verb in our sense:
liegen
O s
auf,unter L
...
where liegen inplies 'horizontal position1. HuS include in V2 such sentences as: (212)
Die Ehegatten liegen in Scheidung. 'The couple is in the process of getting a divorce. 1
(213)
Das Institut liegt an der Spitze.
(HuS 1973:359)
'The institute is in a leading position. 1
105
It is obvious that in these exanples liegen is not to be connected with the L notion; it functions rather as a "function verb1 of the predicate: in Scheidung liegen, an der Spitze liegen.
Here HuS's syntactic valence plus semantic specification are insufficient. In Scheidung and an der· Spitze are not independent actants but form the predicate together with liegen. (214)
In a sentence like:
Muß ich liegen? 'Must I stay in bed?'
the major communicative function is in the distinction 'to be in a lying position1, the L not being relevant, although it usually implies 'in bed1. It appears that, according to HuS, the verb sitzen implies -KAnimate, not just -ttiuman. According to HuS: HuS
V^
sitzen
sitzen O s
Sn (pS)
} + 2
L
If sitzen is used with 0S only, the sentence-semantic content takes on a very specific reading: (215)
Er sitzt (im
Gefängnis).
'He is in prison.'
In Test Item 1.4, 24 students did not make a distinction between hängen-, and hängen-. 1.4
The context in Test Item 1.4 clearly identified the two verbs. hängen
A:
Ich
d
Mantel an d
haben/sein? B:
Wo
A:
An d
d Tür.
Mantel?
Tür
106
HuS
V
(216)
h ngen
(befestigt
sein)
Sn pS/Sd
Das Bild h ngt an der Wand, 'The picture is hanging on the wall.'
, .. hangen
-—
... hangen.
—
2
A
an, ber . . . ;
a —
an. ber . . .
O
L_.
Dir
For all four verbs stehen, liegen, sitzen, and h ngen, one can use sich befinden or sein if the focus is on L and the distinctions "vertical1 and 'horizontal1 are neutralized. (217)
Das Bush befindet
sich auf dem Tisch.
'The book is on the table.' (218)
Das Kind befindet
sich auf dem Boden.
'The child is on the floor.' (219)
Die Vase befindet
sich am Fenster.
'The vase is by the window.' . , , , . - , sich befinden
Refl
in,auf
O
s
...
L
3.2.3.2 Action-Process/Locative When comparing the O L verbs with closely related Α Ο Ιγ,ίτ
ver
ks, it
is
useful to base the observations on the test examples since these verbs were given in pairs such as 'stehen'/'stelien'. In 1.50, 28 students chose the wrong case for Lp.
('stellen'); in 1.35 ( ' l i e g e n ' / ' l e g e n ' ) , 52 of those
tested either selected the wrong case or did not discriminate between ' liegen'/'legen*.
In 1.37, with 'setzen'/'sitzen',
takes with the L/Lp.
distribution.
56 students made mis-
107
setzen
Sn Sa pS
1 egen ,
Sn Sa pS
V
stellen^
Sn Sa pS
V
h ngen ..
Sn Sa pS
ν
HuS
ι
v
i l
1
This classification is paralleled by our logico-senantic specification __
a __
anfauf ^
...
Dir
(220)
Er stellt die Vase auf den Tisch. 1
(221)
He puts the vase on the table.'
Sie setzt das Kind auf den Stuhl. 'She seats the child on the chair. 1
(222)
Kir legen den Teppich ins
Wohnzimmer.
'We put the rug into the living room.' (223)
Sie h ngt die Tasche an die T r. 'She hangs her pocketbook on the door.'
If Ο is «»referential with A, reflexivization takes place. Ihis can be captured in a general way in our specification, as already shown with A 0 DIR verbs: setzen stellen legen ...
hangen
__ Α
.
a Ο
an,auf i .
Dir
...
_ ~
Refl ~~*"5
These verbs cause imnense problems for the English speaker learning German, partially due to the German case distinction, and partially due to the frequent use of English put for all four German verbs as specified above.
108
Summary of O L and A O L^.
relations:
sitzen liegen O s
stehen
L
hängen
setzen
seat
legen . A stellen l
O J
L
Dir
lay
Refl
O = A
A
hängen
If the I^.
O
Ln. Dir
hang
is not expressed with setzen, and A = O, as in:
(224)
Er setzt sich. 'He takes a seat. '
(225)
Bitte, setzen Sie sich.
(225a)
Bitte, nehmen Sie Platz.
(4.18)
'Please, take a seat.'
the I
is still implied.
HuS have no less than nine 'variants' with stellen; and if we consider prefixation, we arrive at approximately 2O possible verbal concepts with stellen as the core. The English put, incidentally, is even more complex. In 1.24, the students had to supply the correct prefix (' _ stellen') t this resulted in 34 wrong choices. Among the prefixes most frequently selected were an-, ein-, hin- and zu-. stellen
A
O
Ln. Dir
put
abstellen
A
O
(L)
leave
(226)
(park)
Ich stelle meinen Nagen vor Ihr Haus. 'I put my car in front of your house.'
(227)
Darf ich meinen Hagen vor Ihrem Haus abstellen?
(1.24)
'May I leave (park) my car in front of your house?'
109
The specification above shows the significant difference between stellen and abstellen. The prefix ab- implies L in a general sense, which is probably why L is facultative.
The Qiglish park is neutral to this notion, focusing
on the action itself. Stellen and abstellen differ in the L/Ivj- distinction. These are extremely relevant specifications in foreign language teaching; sentence models explain these differences only on the morpho-syntactic level. A similar relation can be observed between holen and abholen in: HuS
holen
/< \-->
(hereinholen)
Sn Sa (pS;
However, such equating of holen and hereinholen is misleading since holen has the facultative Iy . , (Goal-Source-Direction), while hereinholen has incorporated in herein, so that (228)
Er holt die Kinder. 'He gets (picks up) the children.
and (229)
Er holt die Kinder herein. "He brings the children inside. 1
differ in their sentence-semantic content. HuS
abholen2
.. , abholen
—— A
Sn Sa a —— O
(pS)
in, auf ... —— (L·}
_ , , . , fetch (pick
up)
3.2.3.3 Action/Locative and Process/Locative abfahren
1.23
gehen
1.2, 2.3
fliegen
4.6
aussteigen
1.27
zurückfliegen
1.29
abholen
2.6
Test Item 1.23, with abfahren, required the insertion of the correct prefix plus recognition of L . . HUS
a b f a h r e n
S
n
(pS)
110
, , ,
abfahren
von,bei ... —
-^
depart
There seems to be a subtle difference between abfahren and depart in that abfahren does not readily accept ^.^1 and depart is neutral as to S or G. abfahren
A/O
(L)
depart
A/O
(LDirS/G
(L)
The facultative character of L is due to ab- which already implies L. The same is true for L~·
in zui-uak--.
zurückfliegen
— °
A/
fly
back
L
DirG
Abfahren and zurückfliegen are representative of L verbs that allow an Action/Locative as well as a Process/Locative interpretation. gehen definitely require the L Q . Q relation (1.2, 2.3).
Fliegen and
This does not, of
course, apply to fahren and fliegen as in: (230)
Er fliegt
eine Cessna.
'He flies a Cessna.' (231)
Sie fährt einen Porsche. 'She drives a Porsche.'
which are considered separate verbs. In 3.2.2.4 bringen
and bringen^ were discussed.
Holen is considered the
antonym of bringen. Compared with abholen, with focus on L, holen focuses on ^irS/G ' The difficulty i11 ^ distinction between L0irS and L and LQ^J-Q and their raorpho-syntactic realizations was shown in Test Item 2.6, where 25 students selected prepositions and/or cases that did not agree with L, such as zum, aus, nach, ins. Aussteigen is still another instance of 'case incorporation1.
While
steigen has the valence characteristics A 1^. , aussteigen (get off) has incorporated in aus-. steigen 3 (232)
A
This is comparable to English get
£_ . Dj.r
Der Junge steigt über den Zaun. 'The boy climbs over the fence. 1
off.
111
aussteigen (233)
A
(L
.
)
(L)
Mir steigen aus. 'We get o f f . '
(234)
Sie stiegen bei der Oper aus. 'They got off by the opera. 1
Even though the notion K~. _ is included in aussteigen, it is necessary to fully realize the I^v-c Lp^g is used, e.g.: (235)
Itor
Pno~syntact:i-cally ^ German when the facultative
Wir stiegen aus dem Bus aus.
A
L .
A
L
L/-L .ZT O
'We got off the bus.'
(236)
Wir stiegen in Mien aus dem Expre zug aus.
L ,
"We got off the express train in Vienna.'
This leads to frequent mistakes of the following kinds: (237)
*Wir steigen das Flugzeug aus. *Sie stiegen den Bus ein.
Incorporation of L (aussteigen) results in the reduction of the obligatory valence. The same problem could be illustrated and explained with such other Action/locative verbs as einsteigen 'get on', herauskommen ''come out of1, hin berlaufen lrun across1, etc. In summary, the logico-semantic valence concept makes it possible to distinguish between various subtypes of location verbs on a semantic basis. Especially useful is the comparison of such pairs of verbs like stellen/ stehen. Systematic considerations of LAvj- ς (^ΠΊ-Γ^ could lead to a better understanding of the function of prepositions and prefixes. 3.2.4
Observations on reflexivity
3.2.4.1 Various types of reflexivity Terminology referring to reflexivity varies to a considerable extent.
Georg
St tzel (197O) speaks of 'facultative reflexive verbs', 'reflexive verbs' (with obligatory sich), 'semireflexive verbs', and 'irreflexive verbs'.
112
Grebe's Duden Grammatik (1966) distinguishes 'real reflexivity' and 'unreal reflexivity1 (Pseudoreflexivität). Joachim Buscha (1972) made an attempt to arrive at a 'valence-typological' differentiation of reflexive verbs. Buscha's and Stötzel's terminology will be used. A distinction of different types of reflexives in teaching German (and possibly in other languages such as French, Spanish, Russian, Romanian) may prove to be necessary to avoid negative transfer from the native language. This transfer often occurs if a nonreflexive verb in the native language is reflexive in the target language. Distinctions in reflexivity are generally made on the basis of affinity of the reflexive pronoun to the verb. 3.2.4.1.1 Facultative reflexivity In sich schämen 'to be ashamed' there is no variant possible without siah, and the reflexive is part of the verb. Sich waschen, on the other hand, can occur with or without the reflexive: (238)
Ich wasche das A u t o . wash the car.'
(239)
Ich wasche mich. 1
1 wash myself.'
In this case, the reflexive reflects the A-O relation, the action-affected being identical with A. This is already an important consideration. The often used 'Subject-Object identity' concept is avoided here since a semantic differentiation is the goal of this research. Waschen could then be specified as: waschen
——
U
.
= A —»·
"wash 1
J
Whenever reflexive (Refl) is not further specified, it is in the accusative case. This specification could be applied to such other verbs as schneiden 'cut1, verletzen 'hurt1, entschuldigen "excuse1, verändern 'change1, etc. Passivization is, of course, not possible with the reflexive variant. (240)
Ich wasche mich,
(241)
*Ich werde von mir gewaschen.
113
3.2.4.1.2 Obligatory reflexivity In sich schämen no 0 is present in the verbal concept. The reflexive identifies an action that is restricted to the A or E. No substitution of the pronoun is permissible: (242)
*Ich schäme ihn.
nor is passivization possible: (243)
*Ich warde geschämt.
The following examples illustrate that A as well as E are possible for this type of reflexive: sich beeilen (244)
A
'hurry 1
Ich beeile mich. ' I hurry u p . '
sich schämen (245)
Ct
'be ashamed1
Er schämt sich. 'He is ashamed.'
That the reflexive sick in the above exanples is not a proform for a relational constant can be seen in cases where the 'sic/z-verb' can be paraphrased by another verb: (246)
Ich begebe mich zur Bibliothek. gehe 1
1 go to the library.'
. . . , sich begeben
Refl A
in,zu ... L . Di r G
The term 'pseudoreflexive1 seems to be rather fitting for this kind of reflexivity.
114
Buscha (1972:152) observes an interesting difference between sich waschen and sich fürchten by using the conjoining transformation: (247)
Ich wasche mich.
wash myself. 1
(248)
Ich wasche mich und ihn.
wash myself and him.'
(249)
Ich fürchte mich.
am afraid. 1
(250)
Ich fürchte ihn.
am afraid of him. 1
(251)
*Ich fürchte mich und ihn.
(252)
Ich fürchte mich vor ihm.
am afraid of him. 1
(253)
Ich fürchte ihn.
fear him.'
These examples indicate that sich fürchten and fürchten are separate verbal concepts. This can be shown by the following specification: sich furchten , , , (Angst haben)
Refl E
vor — (O)
fürchten
——
--
»aschen
—
-^j-
., ,. . , be afraid
,. of
'fear'
O=A
Refl -- —
L— -J Such important differences as ES 0 versus E 0 and A (0) are not captured by Buscha 's treatment of reflexives. Buscha also does not go beyond the morphosyntactic considerations when he only suggests considering sich etwas vorstellen and jemanden vorstellen as separate verbs and not variants of one verb, vorstellen, (254)
Ich stelle mir sein Erstaunen vor. (can) imagine his astonishment."
(255)
Ich stellte ihm meinen Freund vor. introduced my friend to him. 1
115
The completely different sentence-semantic content underlying these two verbs is evident in: sich vorstellen
Refl
d
O
(DIR )
Ob Ο
imagine
and vorstellen
A
Ο = Ά Ο -~
Refl Ο
Ο = A Ο
Refl Ο
I....J introduce
A
Ob Ο
l
4
Therefore: (256)
Ich habe mich vorgestellt. Ί
(257)
introduced m y s e l f . '
Er hat mich vorgestellt. 'He introduced me. 1
(258)
Darf
ich vorstellen?
Herr
'May I introduce Mr. . . . ? '
but not: (259)
*Ich habe mir vorgestellt.
(260)
*Er hat wir vorgestellt.
The frequency of this type of error, which is due to a vague concept of the semantic sentence structure connected with pairs of verbs such as vorstellen and s-ioh etwas vorstellen, cannot be overlooked by any observant foreign language teacher. Another case is erinnern 'remind1 and sieh erinnern 'remember1: (261)
Ich erinnerte ihn nicht an die Pr fung. Ί did not remind him of the exam.'
(262)
Ich erinnerte mich nicht an das Examen. Ί
did not remember the exam."
116
, , . sich erinnern
erinnern
Refl an,daß,w -- - -
a an/daß — — ---
, remember
. , remind
——
Ob/that, w - --
Ob of '/that, w — — ---
A similar distinction, as Zarechnak (1974:33) points out, appears to be made in Russian. When discussing kaufen/verkaufen in 3.2.2.4, the reflexive was introduced to account for the coreferential relation between A—D I R , as in (263)
Er kaufte sich einen Porsche. 'He bought himself a Porsche.'
Comparing this relationship to the one in a sentence like (264)
Er wäscht sich die Hände. "He washes his hands. 1
an important difference must be noted, namely, that while in (263) it is a coreferentiality relationship, in (264) the sich indicates Ojj^ (the inalienable possession relationship) . Other instances of the OInai relation are: (sieh) die Haare kämmen 'to comb one's hair', (sich) den Mantel anziehen 'to put on one's coat1, etc. Buscha enters these under 'dativus possessivus1 (Buscha 1972:156) . All verbs with the , relation belong to the facultative reflexive group. 3.2.4.1.3
Semireflexivity
This type of ref lexivity requires the reflexive but permits substitution within the pronominal paradigm: (265)
Ich ärgerte ihn mit dieser Sache. irritated him with this matter. 1
(266)
Ich ärgerte mich mit dieser Sache. ' irritated myself with this matter. 1
but (267)
Ich ärgere mich über das schlechte Wetter. ' I am annoyed about the bad weather . '
(268)
*Ich ärgere ihn über das schlechte Wetter. ' I annoy him about the weather . '
117
This semireflexive type corresponds to Buscha's pseudoreflexives (Buscha 1972:157). In 3.2.4.1.2 above, fürchten/ sieh fürchten were discussed and entered as separate verbs. ärgern
The same is suggested for ärgern/ sich ärgern.
_
a __
mit ___
Antäte someone
—
Ob —
with -^j-
L....J . , argern ·. sich
(269)
-- Refl --
über (Q)·
,_ get , .irritated · · , _ _ , to
-
about — —^
Er ärgerte sie mit seinem Benehmen. 'He annoyed her with his behavior."
(270)
*Er ärgerte sich mit seinem Benehmen.
(271)
Er ärgert sich über sein Benehmen. 'He was irritated with his own behavior."
Other verbs of the semireflexive type are:
sich interessieren für "to be
interested in', sich beschäftigen mit 'to be occupied with1 . 3.2.4.2
Special functions of reflexivity
3.2.4.2.1 Proform for passive This use is rather restricted. 1
be referred to as 'reflexive (272.)
It is questionable whether the proform should at all:
Die Tür läßt sich nicht schließen. 'The door does not close (cannot be closed) . '
(272a)
Die Tür kann nicht geschlossen werden.
(272b)
Man kann die Tür nicht schließen.
In German, mem can be used as a substitute in passive sentences with the nonspecified A. (273)
In diesem Sessel kann man bequem sitzen.
(273a)
In diesem Sessel sitzt es sich bequem. 'This chair is comfortable to sit in. 1
118
Without repeating the valence characteristics of such verbs as sitzen, verkaufen, it can be observed that the s iah in the above sentences in no way implies the notion of reflexivity. It is believed that the reflexive concept used in this context is rather confusing for language students.
3.2.4.2.2
Stative passive versus state reflexive
G. Helbig and Fritz Kempter (1973:18) distinguish the following sentences as being derived from different sources: (274)
Das Mädchen ist
verlieht.
'The girl is in love. (275)
= state
reflexive
1
Das Mädchen verliebt sich. 'The girl falls in love.'
ist verliebt**
sich verliehen
(276)
ist
Der Brief
geschrieben.
= Stative passive
'The letter is written.' (277)
Der Brief
ist
geschrieben worden.
'The letter has been written.' ist geschrieben *
ist geschrieben worden
Morpho-syntactically there is no difference between (274) and (276). This fact leads to ambiguity in cases where the verb permits facultative reflexive. Das Kind ist (278)
Das Kind ist
gewaschen worden. (Passive)
/ gewaschen. Das Kind wäscht sich.
(Reflexive)
While the "state reflexive1 indicates the resultative state of A, the 1 stative passive1 refers to the that has been affected by the action. (279)
Das Auto ist
repariert.
O
S
"The car is repaired.' (280)
Das Auto wurde repariert.
O (A)
'The car was (being) repaired.'
The stative passive usually does not allow the A to be expressed.
119
3.2.4.2.3 State reflexive/stative passive and attributive function The extensive use of participial constructions in German, especially in the written language, is extremely troublesome for the learner of German as a foreign language. of concern here.
The potential complexity of such constructions is not What is of importance is the relationship between logico-
semantic valence, the reflexive, and the realization of participial structures. (281)
Das Kind wäscht sich.
(28la)
Das sich waschende Kind ... 'The child washing himself . . . "
The reflexive is included in the participial construction when used attributively.
This process is not to be confused with nominalization, as in:
(282)
Das Waschen des Kindes ... 'The washing of the child ...'
In order to preserve the once established sentence-semantic content when used as part of a more complex sentence-semantic content, the reflexive must be carried over into the participial construction. the sentence semantic content changes, as in: (283)
Die Mutter wäscht. 'Mother is washing (doing the l a u n d r y ) . '
(283a)
Die waschende Mutter
...
"Mother who is washing ...' (284)
Die Mutter wäscht sich. 'Mother is washing herself.'
(284a)
Die sich waschende Mutter ... 'Mother who is washing herself . . . '
The same applies to the fürchten/'sich fürchten pair: (285)
Das Kind fürchtet sich.
• . furchten *·· u.t sich
Refl —
vor
If this is not done,
120
(285a)
Das sich fürchtende
(285b)
*Das fürchtende
Kind ...
Kind ...
The specification for fürchten (E 0) indicates that 0 is obligatory and must therefore be carried over into the participial construction: (286)
Das Kind fürchtet den großen Hund. 'The child is afraid of the big dog.'
(286a)
Das den großen Hund fürchtende kind ... _ 'The child who is afraid of the big dog ...'
It must be noticed, that with obligatory reflexives, e.g. sich verlieben, sich does not occur with the past participle (it has already disappeared with the state reflexive). (287)
Das Mädchen verliebt sich.
(active)
(288)
Das Mädchen ist
(state reflexive)
(288a)
Das verliebte Mädchen ...
verliebt.
Sich must, however, occur in the present participle (287a)
Das sich verliebende Mädchen ... 'The girl who is falling in love ...'
The valence characteristics of the verb are reflected in the participial constructions: (289)
Die alte Frau verkauft
Blumen.
'The old lady is selling flowers. 1 (289a)
Die Blumen verkaufende
(289b)
*Die verkaufende
alte Frau ...
alte Frau ...
In passive sentences we have valence reduction due to the facultative use of A.
Therefore the passive is not representative of the valence of the verb. (290)
Die Blumen sind (vor der alten Frau) verkauft
(29Oa)
Die (von der alten Frau) verkauften 'the sold flowers . . . "
Blumen
...
worden.
121
The importance of the valence characteristics for participial constructions can also be shown with other verb types such as: sich befinden (291)
OL
Herr Huber befindet
sich in der Bibliothek.
'Mr. Huber is in the library. 1 (291a)
Der sich in der Bibliothek befindliche
Harr Huber ...
'Mr. Huber who is in the library . . . " (291b)
*Der sich befindliche
(291c)
*Der in der Bibliothek befindliche
3.2.4.3
Herr Huber ... Herr Huber ...
Test items with the reflexivity function
After this brief survey of reflexivity, which constitutes a very difficult matter in foreign language learning, various test items will be considered in an attempt to evaluate the usefulness of valence specifications. 3.2.4.3.1 Ausziehen 'to take off clothes', 'undress1 The students had to supply the correct prefix in Test Item 1.4O/ which resulted in such erroneous choices as auf-, ab-, an-, un-. HuS
ausziehen..,
(auskleiden)
Sn Sa
Sa = Sn
»-(Reflexiv) Ο = A
ausziehen
(292)
O
(DIR) G
DIR^ = A G
Ich ziehe ihm die Schuhe aus.
Ich ziehe die Schuhe aus. Ί
take off my shoes.
1
DIR
A O DIR
'I take off his shoes.' (293)
O
AO
G
G
~-Refl *~Refl
d
122
(294)
Ich ziehe mir die Schuhe aus.
A O DIR
take off ray shoes.' (295)
Ich ziehe mich aus.
AO
'I undress myself. 1
The following distinction is not realized in English: (296)
Er zieht ihm die Schuhe aus.
A O DIR„
(297)
Er zieht sich die Schuhe aus.
A O DIR
G
A comparison with English undress/take off clothes complicates the description even further: undress
A
(DIR„)
DIR_ = A
G
DIR
G
(298)
She undresses the child.
(299)
She undresses
take off (300)
(301)
(clothes)
G
*~Refl
'ausziehen'
(herself). A
O DIR
G
I take off
my shoes.
'Ich ziehe
(mir) die Schuhe aus. 1
'ausziehen'
I take off his shoes. 'Ich ziehe ihm die Schuhe aus. 1
In spite of the complexity of morpho-syntactic realizations, it is possible to make use of the same relational constants for the specification of 'undress '/'take off' and 'ausziehen'. The same descriptive pattern could be applied to verbs such as anziehen 'get dressed1,'put on', and aufsetzen 'put on a cap','hat'. 3.2.4.3.2 Ansehen'to look at" Test Item 4.24 reads: Karl watched the entire soccer game. 'Karl hat sich das ganze Fußballspiel angesehen.'
123
In the students' responses, besides ansehen, the following verbs were used: beobachten, schauen, anschauen, betrachten, and sehen: HuS
V
ansehen
ansehen, (anschauen)
look at
(anblicken, anschauen)
_ A t
_a_ O l
= A O —~Refl
Ob ——
A
=A
C/
* (302)
o
Sn Sa
»-Refl
Er sah seinen Freund an. 'He looked at his friend. 1
(303)
Er sah sich an. 'He looked at himself. 1
HuS
V
ansehen
(304)
(intensiver betrachten)
Ich sah wir das Fußballspiel ' I watched the soccer game.
(305)
Sn Sa
an. 1
Er sah sich die ganze Vorstellung an. 'He watched the entire performance.' Refl
ansehen2
d
a —
vatch
—
O b —
In order to get a more detailed picture of the vrord-field sehen, several other related verbs are analyzed and oonpared: HuS
V.
sehen- /i\-?
(wahrnehmen, blicken) Sn Sa/pS/NS
(I) standing for Infinitive, as in: (306)
Ich sah ihn kommen. 'I saw him come.' a/daß,ob,w
sehen
—
(307)
Ich sah, daß er keine Zeit hatte. saw that he had no time. 1
daß,ob,w
(I)
124
The (I) of HuS is not considered since it represents a separate sentencesemantic content that is in a dependency relation to O. The important difference between sehen and anschauen is the Ε Ο/Α Ο specification. (not
zusehen entered in HuS)
A
watch, look on
—
d (DIRJ
(O)
G
-^-
(DfR
O + DJ«G
}
O—
Inf
G
(308)
Wir sahen bei der Probe zu. 'We watched the rehearsal.'
The 0 can be omitted but is always conceptually present. The 'meaning' contribution of zu- is 'watch an activity for some time', i.e. focus on activity and time: (309)
Wir sahen ihnen beim Tennisspielen 'We watched them play tennis.
zu.
1
This activity orientation is evident in the English infinitive play, HuS
beobachten
, . beobachten (310)
Sn Sa/NS da ,ob,w
a/da ,w —
—— Λ
. observe
Ο
—τ— A
Ob/that/w — \J
Mir beobachten den Vogel. 'We observe (watch) the bird.'
(311)
Er beobachtete, wie die Frau die Stra e
berquerte.
'He observed how the lady crossed the street."
If beobachten is compared with zusehen, it can be observed that beobachten implies 0-focus, while zusehen focuses on the activity. HuS
V
betrachten
betrachten (312)
——A
—— C/
(lange ansehen)
Sn Sa
view, look at
—τ— A
Der Besucher betrachtete das Gem lde. 'The visitor viewed the painting. 1
—τ— C/
125
With betrachten the time element is stressed. HuS
besichtigen-
besichtigen (313)
Sn Sa
A O
view (inspect, visit)
AO
Sie besichtigten die China-Ausstellung. 'They visited the China exhibit.'
(314)
Der Student besichtigte die Wohnung. 'The student inspected the apartment.'
Having concluded our illustrative investigation into see/look, we can summarize our observations as follows:
The logico-semantic valence specifi-
cation provides for such basic distinctions as A 0 versus E 0.
It makes
comparisons between languages possible by using the same relational constants, and it indicates the language-specific morpho-syntactic realizations.
It does not provide for detailed semantic specification of Ο which
is decisive in the distinction between zusehen, beobachten, and besichtiget:. It nevertheless is an improvement over HuS's entries or entries in commonly used dictionaries. sehen
E
see
Ο
L___J
ansehen.
look at
ans en en y
Λ— rtexj. u ———— —
WdL-i.Il
zusehen
A (DIR ) Ο
watch, look on
„
focus
beobachten focus
—
betrachten focus
—
besichtigen focus
(0)
'activity'
Ο
O
Ο + 'time (duration)'
A
O
(O)
(DIR )
observe, watch
A
A
Ο
O
'spatial'
view, look at
look at, inspect, visit
A O (for some time)
126
One of the major problems encountered by the English speaker in learning these German verbs is the wide range of meaning of look (at) and watch. The uncertainty in this respect was proven by their responses in Test Item 4.24. One way to remedy this situation is the incorporation of aspects of logioo-semantic valence into a cyclic approach for the treatment of verbs (predicates) . The concept of such an approach will be illustrated in 4.2. 3.2.4.3.3 Ändern "to change1 Test Item 4.24: He has changed. 'Er hat sich geändert. ' ( " E r ist anders geworden. 1 )
The following verbs were mistakenly used:
wechseln, verwechseln, ver-
wandeln, umziehen, and wenden. HuS
ändern
- ,, (ver), ändern (315)
Sn Sa
Refl
-——
change
Das Wetter (ver) ändert sich. 'The weather changes. 1
(316)
Die Zeiten ändern sich. 'Times change. '
andern2
—
—
change
L _____ l (317)
Er ändert seine Haltung. 'He changes his attitude. '
(318)
Sie hat sich geändert. 'She has changed. 1
(319)
Wir ändern unsere Pläne. 1
We change our plans . '
—
——
127
Thirty-one students did not use the reflexive as required in 4.34.
Test
Item 4.23 was: The weather changes often here. 'Das Wetter ändert sich hier o f t . ' 29 students emitted the reflexive.
3.2.4.3.4
Further problems with reflexivity
2. 7
anmelden
'register'
AO
'apologize'
AO
21 students omitted reflexive; 4.27
entschuldigen
32 students omitted reflexive. In 3.2.4.1.2 erinnern/'sich erinnern was discussed.
A related verb of
thinking was presented in Test Item 4.5: (32O)
Please remember that there is no lecture tomorrow. 'Bitte denke daran, daß morgen keine Vorlesung ist."
Thirty-one students used (sieh) erinnern (an)', 12 omitted the reflexive; 24 the preposition an. HuS
V
denken (321)
denken an
——
an/daß
Sn pS/NS
daß
L . , think
——
Ich dachte nicht an das Examen. did not think of the exam. 1
(322)
Erinnerst du dich an das Examen? 'Do you recall the exam? 1
(323)
Ich erinnerte ihn an das Examen. "I reminded him of the exam. 1
of/that —-
128
In HuS there are no less than five variants for denken. the same syntactic valence, namely 2. HuS
V
(324)
sich denken-, (sich vorstellen)
All but five have
Sn Sa
Adj/pS/Adv
Er denkt sich die Sache leicht. 'He imagines the matter to be easy. 1
denken (sich vorstellen)
(325)
Refld E
a O
j
Adj/Adv/als O„ ,
I Mod
Ich denke mir das Examen leicht. 'I imagine the exam to be easy.'
(326)
Sie denkt sich die Ferien anders. "She imagines the vacation to be d i f f e r e n t . 1
think (imagine)
E
Ob O |
Inf+Adv C> , | Mod
This verb is an example of a modification (modal) as part of the obligatory valence. 4.41
verletzen/sich
verletzen
Fortunately nobody was injured. 'Zum Glück war niemand verletzt. 1
The following other verbs were used in the test (all within the conceptual field 'something negative happening to someone/something'): beschädigen, zerstören, brechen, schaden, passieren, wehtun. Evidently the students were scanning the specific conceptual field, but had not yet any firmly established associations with a particular verb corresponding to injure. Dictionaries are of little help. Cassell's (1962) has the following for the entry schaden—harm, hurt, injure, damage. verletzen. 1 (physical) to inj ure
—
a —
a -^j-
an -jj-j-
E = A E = A + O =A +L
E »- Refl E —^ Refl d E »- Refl
129 (327)
Ich verletze 'I
(328)
ihn.
injure h i m . '
A
Ich verletzte mich. " I injure m y s e l f . 1
(329)
E = A
Ich verletzte mir die Hand. Ί
(330)
injure my hand. 1
E = A + Ο
Ich verletzte mich an der Hand. ' I injured myself on the hand.'
verletzen2 (psychological)
3.2.5
E
A
a E
hurt
Ε = A + L
(someone's feelings)
On reciprocal verbs
There is a certain ambiguity between reciprocal and reflexive function in German because the plural reflexive forms are identical to the reciprocal. Mir waschen uns. (Mir
We wash ourselves. We wash each other.
waschen einander.)
As with the reflexive verbs, there are some reciprocal verbs that are obligatorily reciprocal, such as sich einigen 'to come to an agreement1, sich anfreunden 'to become friends', siah verabreden 'to make an appointment'; others take the reciprocal function facultatively, such as: sich verheiraten "to get married1, sich treffen 'to meet'. Test Item 4.19 dealt with sich treffen ' to meet'. They met in front of the library 'Sie trafen sich vor der Bibliothek. 1
Disregarding mistakes with locative and the use of related verbs, 36 students did not use the reciprocal sich. HuS
V
treffen
(begegnen)
Sn Sa
130
treffen. 1
a —— Ε
—— Ε
Ich traf ihn.
2
(331)
.
—— Ε
Ob —-— Ε
'l met him.'
If E- plural = E 2 , then E2 ^treffen,
meet .,
.,
(by coincidence)
A
*- Reciprocal
&
meet
E
(by arrangement)
Wir trafen sie
Ob A
E
um zwei Uhr.
'We met them at two o'clock.'
3.2.6
Verbs implying time relations
In 2.2.4.4.5 the time relation was briefly discussed. small group of verbs that require "time content.
1
There is a relatively
as part of the sentence-semantic
In Test Item 2.15, the verb stattfinden
'to take place1 was given.
The major problem was the assigning of the correct tense to the time-related verb stattfinden
(19 errors).
... , stattfinden
—
am, .../Adv
bei, vor .../Adv
The "durative1 time relation as in dauern was included in Test Item 1.8.: dauern
_ o
a/Adv
r_ Dur
.In the question, bis 'until' was to be supplied; 32 students either omitted bis or used another word. 3.2.7
On predicates of the type 'be hungry1, 'get tired of
Looking back on bekommen in 3.2.2.2, there was one specification for bekommen belonging in the Ε Ο group: (166) (166a)
"1,
Ich bekomme Hunger.
-)
Ich werde hungrig.
(
am getting hungry.'
131
Test Item 4.2 was: am getting hungry. 19 students answered with a be or have form: Ich habe Hunger (Ich bin hungrig). Other incorrect renderings for getting hungry were: *Mir wird hungrig. *Es ist
mir hungrig werden.
*Ich habe Hunger bei mir.
The difference between hungrig sein and hungrig werden could be stated as: (332)
Ich bin hungrig.
(332a)
Ich habe Hunger.
E
'I am hungry.' (166a)
ich werde hungrig.
(166)
ich bekomme Hunger.
E
'I am getting hungry.'
The 'noun—have1 type is not possible in English. Zarechnak (1974:34) contrasts examples like: (333)
I was satisfied.
E
'Ich war z u f r i e d e n . " (334)
I felt
satisfied.
'Ich fühlte mich zufrieden.
A-E 1
The A-E type reflects the "subjective" notion, since the emotion, or psychological state, is not a result of outside forces.
Considering some of the
incorrect responses above,, it seems that the students followed the -=— fc-s concept with *Mir wird hungrig, which is not possible with predicates referring to sensations relating to food (drink). (335)
Mir ist kalt. 'I am cold.'
(336)
Mir ist 1
nicht gut.
1 do not feel well.'
Observe, however:
132
Returning to the notion +subjective] as observed above and also stated by Rüdiger Zimmerman (1972a:238), the following distinctions can be made: (337)
Er war
zufrieden.
-subjective
'He was satisfied. 1 (338)
s
Er fühlte sich zufrieden.
Refl
'He felt satisfied.' (339)
, . .. +subjective
s
Er wurde zufrieden.
. +inchoative
'He became satisfied." (340)
I (my answer) satisfied
him.
Ich (meine Antwort) befriedigte ihn.
a A/O
Test Item 4.36: I'm getting tired of him. "Ich werde seiner m ü d e . 1
The variety of erroneous responses was amazing, partially caused by the addition of him. Some examples: *Ich bin müde von ihm. *Ich müde mich an ihm. *Er freut mich nicht mehr. *Ich werde an ihm ermüden.
The students searched for paraphrases in English and then tried to put those into German.
The high degree of difficulty of short sentences such as the
one in 4.36 is in part due to the unsystematic treatment of predicates which, on the morpho-syntactic level, are not realized as full verbs. Do adjectives have valence characteristics, like verbs?
Karl Ernst
Sommerfeldt (1971:115), when discussing the valence of adjectives, refers to adjectives that can require more than one relational constant as 'relative adjectives' (see also Sonmerfeldt/Schreiber 1974). 1
an instance of such a "relative adjective . müde sein
—-— E s
-**— O
müde werden
Müde, in 4.36, is
133
This specification could be applied to other predicates such as fähig sein "to be capable of, würdig sein "to be worthy', ledig sein "to be free All of these 'relative adjectives' change their meaning when the 1
omitted.
This has to do with the general fact that 'two-place
refer to relationships between entities such as (341)
of.
is adjectives
.
He is worthy of the Nobel Prize. 'Er ist des Nobelpreises würdig.'
while 'one-place' adjectives specify qualities (size, form, taste, color, psychological states, etc.) of one entity.
Therefore, we consider würdig
sein and einer· Sache würdig sein to be instances of separate sentence—semantic contents. (342)
Er ist
ledig.
'He is single.' (343)
Sie ist
s
aller Pflichten
ledig.
g
'She is free (exempt) of all her duties."
a
These two-place predicates can have varying underlying sentence-semantic contents. . . . . lieb sein
——
d ——
, , be dear
——
to ——
gut sein
—
zu ——
, , be good
—
to —^—
Catparing Das Essen war gut with Er war gut zum Kind reveals what different sentence-semantic contents underlie the above sentences, differences not captured by 'sentence models'.
interessiert sein an
A
In:
„,, O
, . , . to be interested in
A
O
134
the second obligatory actant is realized in the form of a prepositional phrase. Sometimes one sentence-senantic content can be expressed by either a verb plus actant(s) or adjective plus actant(s): (344)
Er ähnelt seinem Vater. 'He resembles his f a t h e r .
(345)
Er ist
O 1
seinem Vater ähnlich.
Res
O O Res
"He is like (similar to) his father.'
We maintain that the inclusion of predicate adjectives into the logicosemantic valence specification goes beyond such distinctions as 'transitive'/ 'intransitive1 adjectives (Jacobs and Rosenbaum 1968). At the same time, it must be stated that a finer semantic specification of groups of predicate adjectives will only be possible with the introduction of features like [+subjective , l+physicall , l+externalj , etc. 3.2.8
Some thoughts on verbal prefixes like German be-, er-, ver-, zer-, and the holistic/partitive/ resultative distinction
Test Item 4.4O was: Years ago he climbed many mountains. 'Vor Jahren bestieg er viele Berge. 1 or
'Vor Jahren stieg er auf viele Berge.'
Mistakes that had to do with the time expression "years ago1 and mistakes with gender are not of concern here. The relation 'climb—mountain" is the main object of interest. Thirty-eight students used steigen alone or with prefixes such as auf-, aus-, ein-, um-, über-. None of those tested used besteigen. There existed almost total confusion on hew to deal with the 'climb—mountain" relation.
A (346)
auf,über ... L . Dir
climb (up,down) *
Er steigt auf den Berg. 'He climbs up the mountain.'
A
up,over ... L Dir
135
The above specification indicates the necessary prepositional phrase. Steigen does not imply completion of the activity. climb '^Dir (347)
Sie bestiegen den Berg in zwei Stunden. "They climbed the mountain in two hours. 1
Besteigen indicates that the climbing activity was completed and the goal reached (in a general sense, the notion 'holistic1 is applicable here). The well-known examples: (348)
Bees are swarming in the garden. 'Bienen schwärmen im Garten.'
(349)
A (L)
The garden is swarming with (teeming with, full of) bees. 'Der Garten wimmelt vor Bienen.'
O L
are instances of entirely different semantic sentence structures and should not be used as prototypes of the holistic/partitive distinction. It just so happens that in English swarm can have these different interpretations. Sane examples from the Action-Process/Locative group: (350)
Sie pflanzt
Blumen (im Garten).
'She is planting flowers (in the garden).' (351)
Sie bepflanzt
den Garten (mit Blumen).
"She plants the garden (with f l o w e r s ) . ' (352)
Das Kind streut Zucker auf den Kuchen. 'The child sprinkles sugar on the cake.'
(353)
Das Kind bestreut den Kuchen mit Zucker. "The child sprinkles the cake with sugar. 1
(354)
Die Demonstranten schmierten rote Farbe an die
Wand.
'The demonstrators smeared red paint on the wall. 1 (355)
Die Demonstranten beschmierten die Hand mit roter Farbe. 'The demonstrators smeared the wall with red paint.'
136
(356)
Der Student klebt Bilder an die Decke. 'The student tapes pictures on the ceiling."
(357)
Der Student beklebt die Decke mit Bildern. 'The student tapes the ceiling with pictures. 1
(358)
Inge goß Masser auf die Blumen. 'Inge sprinkled water on the flowers.'
(359)
Inge begoß die Blumen mit Wasser. 'Inge sprinkled the flowers with water.'
A general observation made with the examples above is the holistic/partitive distinction. partitive
~Ä~
a ~Ö~
an,auf ... (L . ) ~ Dir
holistic
~Ä~
a ~L~
partitive/holistic mit (O)
pflanzen
bepflanzen
'plant'
streuen
iestreuen
"sprinkle 1
schmieren
beschmieren
'smear'
kleben
bekleben
'tape ( g l u e ) 1
gießen
begießen
'water'
As far as the German examples are concerned, the A 0 (L-. ) group focuses on 0, the A L (O) group on L. way.
The English group is interpretable in either
The 'with-phrase' in the sentences above is not Instrument but Object.
Luise Pusch (1972) mentions that, in German, verbs derived from nouns by using the be- prefix usually have the holistic interpretation, such as bekleiden 'dress', bedachen 'roof, besohlen 'sole', bemannen "man 1 , etc. If the Instrument-Tool relationship is used, the 'holistic1 interpretation is not possible: (360)
*Er begoß die Blumen mit einer Gießkanne. 'He watered the flowers with a watering c a n . '
The relative closeness of the Object to the semantic sentence structure, compared to the Instrument, is evident in (361)
Sie beschmierte die Wand (mit Farbe).
(362)
*Sie beschmierte die Hand mit einer Sprühdose. 'She smeared the wall with a spray can. 1
137
If an additional relation is expressed, one usually expects the Object rather than the Instrument. Obviously the introduction of the prefix bewith the German schmieren resulted in a change of semantic sentence structure with a simultaneous shift of focus. The verbal prefixes be-, er--, ver-, zer-, in German, fulfill lexico-semantic functions in that they carry the notion 'holistic' with such further specifications as location (be-), resultative (er-), or nihilative (zer-). John Platt (1971:41) considers such notions 'deep cases' (see also Günther, 1974). This view is not shared in this thesis. They are rather seen as focusing devices with specific semantic values as specified above. (363)
Sie pflanzte
Blumen (hinter dem Haus).
'She planted flowers (behind the house).
In these sentences the "normal1 semantic relation between the verbal concept pflanzen and the O in Blumen is present.
This is not the case in the
English: (364)
She planted the garden with flowers.
since one does not "plant gardens'. In German, this change of the semantic sentence structure and focus is realized by the prefix be-: (365)
Sie bepflanzte
den Garten (mit Blumen).
This is why bepflanzen is specified as A L (0), while the English "plant1 in this sense requires A L O (the comnunicative situation may, of course, make the omission of possible). Test Item 4.37 was: Do you know who broke the vase? "Weißt du, wer die Vase zerbrochen hat?"
44 students used brechen, which resulted in such sentences as: * ... , wer die Vase brach? * ... , wer die Vase gebrochen hat?
138
This indicates that the notion of zer- (nihilative), which is obligatory with brittle objects like glass, was not understood.
It must be stressed
that these 44 students knew the German verb for 'break1. HuS
V
(366)
zerbrechen
(zerschlagen I
Er zerbrach die
Sn Sa
Tasse.
'He broke the cup. 1
HuS give six variants for brechen and throe for zerbrechen. zerbrachen , +nihilative\
(a) (A)
o
'break (to pieces)'
Other verbal concepts fitting this description are: zerbeißen
"crunch, bite through 1
zerdrücken
'crush, squash1
zerfleischen
'lacerate'
zerfressen
'eat, gnaw away 1
zerlegen
'dissect, decompose'
zerreißen
'tear to pieces, rip up 1
zerschneiden
"cut to pieces'
zerstören
'destroy'
It must be stated that
er- does not automatically imply an A in the seman-
tic sentence structure (serbrechen) ·, on the other hand, many verbs require A. Although many of the basic verbs above are used in textbooks, their zer- correlates are hardly ever mentioned. Little help is provided if only once throughout the language program the general notion of z er- is explained. Only by a systematic and repeated treatment and practice of such prefixes as zer- will the students achieve a certain competence in using them.
139
Test Item 4.3: She broke her arm. 'Sie brach (sich) den Arm. 1
brechen
—
—
DIR =
DIR
«- fief I d
i 16 students failed to use the reflexive; 31 translated the English possessive which is rendered in German by Ref1 (d). Contrasts must not be underestimated in the discussion of such verbal concepts as brechen, sich brechen, zerbrechen: (367)
Sie brach sich den Arm.
(368)
Er zerbrach die Vase.
(369)
Er zerbrach an seinen Sorgen. 'He despaired of his sorrows.'
Also: (370)
Er zerbrach sich den
Kopf.
'He racked his brains. 1
not
*'He smashed his head.' Only if the different possibilities from concrete use to abstract to idiomatic are brought into focus can the student at a more advanced stage enjoy the richness of the ways concepts can be realized, and at the same time achieve a competence that reaches beyond sentence patterns. With er- the notion of "getting into a certain final (negative) state1 is often present: schi eßen
' shoot '
erschießen
'shoot to death"
trinken
•drink'
ertrinken
1
erbleichen
'become (turn) pale*
ermüden
'get tired'
bleich sein
'be pale
1 1
drown'
müde sein
•be tired
schlagen
'beat'
erschlagen
•beat to death1
rot sein
'be red'
erröten
'blush 1
wach sein
'be awake"
erwachen
1
awaken'
140
The er- verbs imply the UresultativeI.
In English this has to be
specified with several verbs (' ... to death"). friere/)
E
erfrieren
E
+resultative
With erschießen some interesting observations can be made: erschießen
A
E
shoot to death
A
E
-fresul ta tive
· schießen
——
—-—
u shoot
In connection with hunting (shoot a deer), even though 'deer1 is animate, schießen is used. (371)
Er schoß ein Reh (einen Hasen). 'He shot a deer (a r a b b i t ) . 1
(372)
*Er schoß seinen Freund.
(373)
Er schoß auf seinen Freund. "He shot at his f r i e n d . "
(374)
Er erschoß seinen Freund. "He fatally shot his friend.
Knowledge of semantic congruence and constraints is important for the use and understanding of language. The example above is evidence of how these constraints can vary from language to language. Ernst Leisi (1967:68) investigated this matter in detail. 3.2.9
A new look at 'function verb constructs1 (Funktionsverbgefüge) of the type to come to conclusion—relevance of logicosemantic valence
In considering foreign language materials that are geared towards the advanced learner, one all too often observes that systematic treatment of logico-seirantic relations is missing. Instead, the focus is on extensive vocabulary expansion. Due to the variety of readings, an almost overpowering amount of individual vocabulary items is to be mastered. "Function
141
verb constructs1 (hereafter referred to as FVG) are introduced as 'idiomlike1 structures of idiosyncratic behavior.
The long-held contention in German
that such structures should best be avoided (often even marked as undesirable stylistic variants of full verbs) can no longer be tolerated in any comprehensive program of German as a foreign language.
Contemporary prose,
newspapers, scientific tests, etc. provide us with frequently used constructions of this type.
It is fair to say that FVG take a medial position
on the axis of idiomatization, hence their frequent treatment as idioms. In defining FVG we follow Engelen (1968) who considers structures of the following type: (Function) verb + preposition + noun (with or without article, sometimes the enclitic form of the article and preposition)
Example:
put into motion^ reduce someone to despair.
We observed in 2.1.3.2 that verbs of cannon usage and seemingly simple lexical structure frequently are given very complex descriptions in HuS's dictionary.
Several of these verbs can be used as the 'function verb1
(Funktionsverb) of a FVG, their 'semantic load1 having been reduced to almost zero.
They may also have been stripped of the valence which they
had as a full verb.
The 'semantic load1 has been shifted to the noun.
It
is not surprising that these 'function verbs' are compared to modals. To identify FVG, Engelen has put together nine criteria. these criteria valid and sufficient.
We consider
EVG being investigated in this section
are, therefore, taken from Engelen without further testing.
Helbig/Buscha
(1972) discuss 'function verbs' and verbal structures with accusative nouns from a syntactic point of view, but in no way as comprehensively as Engelen. Wolfgang Herrlitz (1973) is the first one to attempt the description of FVG in a generative transformational framework. Are FVG synonyms of full verbs? This point of view is often evident from dictionary entries, or for that matter, in Helbig/Buscha (1972:75). pare, for instance: in Bewegung bringen
= bewegen
'to set in motion 1
zur Aufführung
= aufführen
'to perform'
= schließen
'to bring to a close"
bringen
zum Schluß bringen
Com-
142
The most frequently used "function verbs' in German are bringen and kommen, as in in Bewegung kommen
'to start moving'
zur Aufführung
'to be performed 1
kommen
zum Schluß kommen
'to come to an end'
Qxjelen attributes to 'bringen1 the function 'transitive1, to 'kommen' the function 'intransitive1.
His main goal is to systematize the function of
"function verbs' in connection with prepositions. It seems that the causative functions, as discussed earlier in combination with modes of action (Aktionsarten) such as 'inchoative', 'resultative', 'durative1, and 'perfective* prove useful in this context. The difference between in Bewegung bringen and
in Bewegung kommen
is that bringen carries with it the notion of causality (in the form of an agent) with the additional feature 'inchoative'.
This notion is absent from
kommen in in Bewegung kommen The following distinction is suggested: in Bewegung bringen
A
O
-/-causative
in Bewegung kommen
O +inchoative
sich bewegen
O
+durative
The 'bringen'/'kommen* distinction not only holds true for FVG referring to physical movement, but also to FVG expressing social environment/status and emotions. in Mode .bringen
'to bring into fashion 1
A
O
1
in Mode kommen
'to come into fashion
in Mode sein
'to be in fashion 1
in Mode bleiben
'to remain in fashion'
Os 0
+durative
143
Also:
in Kontakt bringen (kommen, treten, stehen, sein, bleiben) ; in Ordnung
bringen, ins Gespräch kommen'to bring in contact with', 'to bring in order1, 'to get into a conversation'.
FVG relating to "emotions' show an interesting
distinction between kommen and geraten, in that kommen implies [+intent , geraten [-intent . in Wut bringen
A E (Action/Experiencer)
'to enrage 1
in Wut kommen
E Uintent (Process/Experiencer)
'to become enraged'
in Wut geraten
E -intent (Process/Experiencer)
'to f l y into a rage'
in Wut sein
Es (State/Experiencer)
'to be enraged'
The +intent is restricted to very few FVG like in Stimmung kommen, in Extase kommen1to get into the mood1, to get into ecstasy, indicating that with Expsriencer verbs it is usually an outside force that induces (motivates) the experience.
This agrees with our general definition of the Experiencer
relation. Another group of FVG uses the preposition uncer and the 'function verbs' stellen/stehen.
All examples are from the terminology of administration
and the courts: unter Aufsicht
stellen
unter Aufsicht stehen
A E E
'to put under supervision 1 'to be under supervision'
The following examples illustrate how the valence of a full verb changes if reduced to a 'function verb': (375)
Der Polizist s t el lte den Verdächtigten unter A
Aufsicht.
'The policeman put the suspect under supervision.
(376)
Der PoJizist ste-Zlte den Wagen unter einen Baum. A
L
DirG
'The policeman put his car under a tree.'
144
The distribution of 'function verbs' in German as illustrated below indicates the systematic behavior of this relatively small group. The diagram at the end of 3.2.9 (p. 146) serves to show, in a model-like manner, general underlying principles; but is not a complete listing.
(For a detailed
semantic description of function verb constructs, see Fink, 1977.)
Such
distinctions, as outlined in the diagram, are of value in a pedagogical grammar.
Putting aside such idiosyncratic relational restrictions as exist
between prepositions and nominal content, the relatively closed system of 1
function verbs' and .prepositions within ' conceptual fields' such as emotion, physical motion, etc. lend themselves to various tasks. The different functions can be illustrated as follows: (377)
Sie setzen das Gesetz In Kraft. 'They put the law into e f f e c t . '
(378)
Das Gesetz trat in Kraft. 'The law went into e f f e c t .
(379)
Das Gesetz ist
O 1
in Kraft.
O
'The law is in e f f e c t . ' (380)
Das Gesetz bleibt in Kraft. ,m, , _ , cc The law remains in effect.
O (durative)
(381)
Das Gesetz tritt außer Kraft.
O
'The law becomes ineffective.' (382)
Sie setzen das Gesetz außer Kraft.
A ü
'They annul the law.'
The preposition außer triggers a reading opposite to i-n in the semantic sentence structure. In many paradigms of this type there will be gaps both in the axis from causative to durative and in corresponding semantic sentence structures in the other language. Another possibility is to consider 'blocks' of FVGs relating to the same 'conceptual field', such as "speaking and discussion': zum Vortrag kommen (gelangen)—zum Vortrag bringen to be recited zur Sprache kommen come up for discussion
to recite zur Sprache bringen bring up for discussion
145
Since the nominal part of the EVG is similar in character to a separable prefix in German, there are structures such as the following: (383)
Zuerst kamen drei Gedichte von Goethe zum Vor t rag. 'First, three poems by Goethe were recited. 1
Unless the student is familiar with the structure of FVG and the different functions of 'function verbs', sentences of this type cause serious difficulties (especially for the listener as against the reader).
The following
contrast illustrates the difference between the full verb and the "function verb1. (383)
Zuerst kamen drei Gedichte von Goethe zum Vortrag.
(384)
Zuerst kamen drei Studenten von Fischer zum Vortrag.
A
L . Dir G
"First, three of Fischer's students came to the lecture.'
Then we reduce the sentences to their obligatory valence: (383a)
Drei Gedichte kamen zum Vortrag.
(384a)
Drei Studenten kamen (zum Vortrag).
Finally, the semantic sentence structure of several FVGs with the same valence can be contrasted: in Verlegenheit bringen
A E
'to make someone embarassed' in Anspruch nehmen "put forward a claim to
A O 1
zur Kenntnis nehmen
E O
'take cognizance of
Illustrations of this kind can help to establish a cognitive awareness of the systematic character of FVGs, and of the problems involved in recognizing FVG in texts and in spoken language.
Φ •H 4-1
Η.
(0
-f.
M
Φ •H HJ
C •r-l
-φ-
^^
(0 M
+
+
Q
O)
c φ
-H
S
rH
-P
4-1 VH
+
+
fl
ω
+
G O •iH 4J
rj
ω φ
C
ο •Η
B;
4->
rB
C
Q)
Φ
•Η 4J U
•H -H
r*
ϊ>
4-) rd 0 U G
m cn Φ DI G
ο
M-l O
in φ
'sS
-H M 4->
ω
•rH Ό -+-
-f-
+
-f
+
+
M •H Φ
M H
1
'SIfl
l
1
Π3 •H
H-> •H C H
-If
-f.
-+-
rH
M Φ U C •H
N
•\_
4-
!D Λ
φ
rH
φ
•H 4J
c
B
0
+
+
+
+
+
+
+
+
t
1
1
Ι
1
Cfl
G (0 H
Φ j>
C
>
C φ
C Φ Φ DI g
(0
Λ Φ
ω
P
D^
C rO •H Φ Di
0 rH Λ Φ
HJ
U
5
1
rH O r4
-P G O
rH
G 0 Φ
rH
c φ ID
4-) -P
(0 G rH
O
(1)
+
4J U)
"
ci φ
c
ft•H
CO
Λ
Φ H
•rH Φ
O •rH •U
υ 3
φ
£> •H 4J
ain
D (0
V
JJ
B
3
-H HJ U)
O •H
G Φ
σ c •H Λ
cφ
cΦ N Φ
to
N -P Φ Ul ^) Φ
C
φ o c^ Φ
C Φ
Λ Φ
ε*
G Φ rH rH φ
4-1 VI
C
φ ^J
rH
147
4.
OUTLOOK
4.1
The psychological validity of the concept of a 'conceptual composition plan' for sentences and discourse
The major purpose of the theoretical part of this thesis was to shew the shift of emphasis from the investigation of primarily morpho-syntactic phenomena to the investigation of semantic (lexico-semantic and logicc—semantic) aspects of language.
The necessity of considering these new insights into
the semantic structure of language in a pedagogical grammar have been discussed elsewhere (1.2).
The analysis of errors, made in the experimental
language test, revealed a variety of difficulties which the learner encounters.
Our area of interest was limited to the investigation of basic semantic
sentence structures (Satzbegriffe) and their realizations in actual sentences. Recent research in psycholinquistics affirms the central iirportance of the logico-semantic composition of the verb (predicate) for sentence and text recognition.
Engelkamp (1973) observed that noun phrases which are obliga-
tory actants of the verb (psychologically implied by the verb) were reproduced more easily than those that are not. true in the other direction:
This observation did not hold
the probability of reproducing the verb was no
greater with obligatory actants than with free adjuncts.
The probability for
reproduction of an entire sentence decreased with the increase in predicateargument relations (Engelkamp 1973:148) .
It is important to know that Engel-
kamp's experiment was intended to verify or disprove the psychological validity of the notion of verb centrality. Another study (Teigeler 1972) reaffirms the validity of the valence concept and the psychological reality of basic sentence models on the content level.
One significant observation in this study is that actants with higher
affinity to the verb were learned more easily than those that were free adjuncts.
Although Teigeler does not deal with the relational constants such
148
as Agent, Object, and Experiencer, it is evident fron examples used in his research that A, O, E are often actants with higher affinity to the verb: Die signifikante Differenz aber zwischen den Behaltenswerten des Substantivs und des Objektivs einerseits und denen der Gliedkerne der freien Satzglieder andererseits spricht deutlich für die mit der Valenztheorie des Verbs erklärten und in der inhaltsbezogenen Grammatik beschriebenen engen Beziehung des Verbs (Prädikats) zu den sogenannten GrundformGliedern und seine größere Distanz zu den sogenannten freien Satzgliedern. (Teigeier 1972:93)
Teigeier's research is of special interest for two reasons: 1)
He takes as a background for his experiment, concepts of grammar as put forward by Brinkmann and Erben (see 2.1.3.1).
2)
His research deals with the relationship between lexical units and the processing of sentences, especially the retention and reproduction of sentences, significant for a pedagogical grammar.
In a study by Kintsch (1974), the difficulty of recall of propositional structures which underly sentences was tested.
This research was based on a
psycholinguistic model in which the representation of meaning in memory closely corresponds to the concept of relational constants as proposed in our study. The interim findings in Kintsch (1974) suggest that notions such as Agent, Experiencer, Benefactive, Location are psychologically valid.
Kintsch also in-
dicates the abstract nature of memory representation but argues against lexical decomposition.
Although it would be premature to draw final conclusions
in regard to the representation of meaning in memory, it is encouraging to see that interdisciplinary research points in the same general direction which was taken in our study. 4.2
The 'cyclic approach1 for the introduction of related verbs (predicates) based on logico-semantic valence
When discussing HuS's valence dictionary in 2.1.3.2, we observed that verbs of common usage often show a very complex description in the dictionary. Many of these verbs are introduced during the first months of foreign language instruction. Such terminology as 'basic meaning1 can be very misleading to the student if he interprets it as 'having the same range of meaning as in (his) native language1. We therefore suggest a cyclic approach for the introduction and gradual expansion of semantic concepts entailed by the verb and other vocabulary items that are closely related.
149
In a model-like fashion this cyclic approach shall be demonstrated with some examples that caused considerable difficulty in the test.
In 3.2.1.1
the following verbs of discourse were analyzed and discussed: fragen
ask
befragen
question, interrogate
eine Frage stellen
put a question
erfragen
inquire into
sich erkundiger.·
inquire
In the textbook DaF 1, on which our verb selection is based, fragen appears in Lesson One in the context: Sie fragen, und wir antworten. 'You ask and we answer.
(DaF 1;11)
1
According to our analysis, fragen was given the following specification: ,. en fra
*
—
a
-W)
nach/ob,w
To)
In Lesson One, fragen was used in a minimal environment.
The next time
fragen appears in DaF 1 is in Lesson Seven, this time in the environment A E; then in Lesson Seventeen, as A E 0; and in Lesson Eighteen again as A E 0.
It can be observed that fragen in DaF 1 is used repeatedly with in-
creasingly complex sentence patterns.
Morpho-syntactic complexity was a
criterion for the use of fragen, and, to a certain extent, a cyclic approach was employed. However, no mention was made of the semantic relations involved. Surely the introduction and consistent use of semantically—based terminology is no more difficult than grammatical generalizations such as 'direct object1 and 'indirect object' which are often elaborated on in great detail. Considering semantic universals simultaneously with the morpho-syntactic specification would provide a much more explicit description of verbal concepts.
Whether such a specification as the one for fragen above appears in
the textbook itself, or is only used as a linguistic input for textbook development, may depend on the goals of the textbook, the age of the learners, and other criteria.
150
It is interesting to note that the alternative Ο as ' occurred in DaF 1. The only related verb used in DaF 1 was befragen in Lesson Fifteen, in a reading section: ... 38% der befragten 1
...
M nner ...
(DaF 1:85)
38% of the men who were questioned . . . '
As was indicated earlier (for details see 3.2.1.1), 39 students in our test failed to make a distinction between fragen and befragen. between fragen/befragen
The contrast
could have been introduced at a relatively early
stage since it does not require any additional granmatica.! knowledge, and it is a distinction so frequent in German that by the end of the introductory level this concept should be very familiar. , fragen
——
befragen
—τ— A
a (E)
—— E
nach/ob,w (O) nach/ob,w (0)
To achieve an understanding of the distinction between the two, fragen as well as befragen have to appear cyclically in ever-increasingly complex sentences, e.g.: cycle 1 (lesson 1)
fragen A
cycle 2 (lesson 2)
fragen A
cycle 3 (lesson 5)
cycle 4 (lesson 9)
E
befragen
A
fragen A E O nach wann
befragen
A
fragen A
befragen
A
E
O
E O nach
cycle 6 (lesson 2 O )
fragen A
E O wo (passive)
E
Ο
ber
ob cycle 5 (lesson 14)
E
befragen
A
E
(passive)
fragen A B O da , . ..
It is understood that aspects of tense, mood, etc. are incorporated into the different cycles. At the end of an introductory course, like one based on DaF 1, a student should not only be familiar with fragen and befragen, but also with related verbs (predicates)» such as: eine Frage stellen
——
d
.
151
fragen um (bitten um) f\
(385)
um
a —-—r—
(-)
\ Ci )
Ich fragte ihn um eine Zigarette. Ί
asked him for a cigarette.'
These related verbs are to be incorporated into the cycles mentioned above and contrasted with morpho-syntactic patterns already learned.
An example
of such an important contrast is: fragen nach/fragen
um (bitten um)
'inquire a b o u t ' / ' a s k for (request) 1
At the intermediate level the continuation of this cyclic approach would include sich erkundigen, anfragen, erfragen, ausfragen,
in Frage stellen, in
Frage kommen. If the implications of the cyclic approach for vocabulary expansion and semantization within the framework of logico-semantic valence are fully realized, then the student will not be confronted with a multitude of unrelated vocabulary items at the advanced level.
He will also be able to
gain insight into processes underlying idiomatization, semantic congruency, metaphor, etc. without being a student of linguistics.
The following phrases
will then be more easily understood: (386)
Das kommt nicht in Frage. 'That is out of the question.'
And matters of noninalization with fragen and befragen could be discussed within the framework of their valence specification. Test items could be selected according to their degree of difficulty based on the complexity of the semantic sentence structure, either as determined in contrast with native language sentences with their entailed semantic adjuncts (in the case of contrastive analysis tasting!, or as determi"«^ merely in comparison with other sentences within the language being taught. In conclusion, it is suggested that the 'syntactic paradigm1, long held to be the organizing principle for foreign language materials, and only recently losing ground to the 'situational paradigm1 (Oiler 1974), should not be replaced by yet another 'paradigm'.
Instead, logico-seniantic information,
152
as made available through research like the one carried out in this thesis, should be allowed to take precedence; and such information, together with the morpho-syntactic possibilities for realization ('syntactic paradigm1) in the specific contexts ('situational paradigm1), should provide valuable information for every aspect of foreign language pedagogy and linguistic investigations.
153
5.
SUMMARY
This study is an investigation into linguistic valence on different levels and its application in a pedagogical grammar.
Theoretical concepts from
current linguistic theories and German grantnars are reviewed and evaluated, and errors made in an experimental verb-centered foreign language test in German are analyzed, providing supportive evidence for the usefulness of logico-semantic valence in describing semantic sentence structures.
To show
similarities and differences between languages (in our case, English and German) in the realization of semantic sentence structures, a combination of logico-semantic and morpho-syntactic valence is suggested. In Chapter 1, the concept 'pedagogical granmar' is defined and the relationships between such a grammar and contemporary linguistic theories are discussed.
It is argued that the terminology used to identify semantic
sentence structures must be universal enough to make comparisons across languages possible. The theoretical background for our study is established in Chapter 2 as follows:
An overview of different valence interpretations is presented,
ranging from Tesniere's concept of valence and the discussion of valence by Erben, Brinkmann and Admoni, to Helbig and Schenkel's valence dictionary, Engel's Satzbaupläne and Vater's generative dependency model.
After some
comments on the relationship of syntax and semantics, criteria are established to identify syntactic units as reflections of semantic concepts. Special attention is given to the distinctions Obligatory/facultative' vs. 'free' in valence theory, and the distinction 'nuclear1 vs. 'peripheral' in case granmar. Sentence semantics and generative semantics provide innovative .insights into logico-semantic aspects of language.
Taking these insights as a point
of departure, relational constants (deep cases) belonging to the causative group are defined by features; the other relational constants basically follow Fillmore's and Chafe's models. cludes:
The list of relational constants in-
Cause, Agent, Force, Instrument; Object, Benefactive, Experiencer,
154
Location, Time, and Final.
These selected relational constants form the
theoretical basis for our model of valence analysis on the logico-semantic level. In Chapter 3, the theoretical considerations regarding logico-semantic and morpho-syntactic valence are applied in the analysis of errors made by university students in the experimental foreign language test in German. The different parts of the test were based on verbs (predicates) taken from Deutsch als Fremdsprache l, a widely used German textbook, and these selected verbs formed the basis for the subsequent analysis of verbal concepts. First, the test objectives, design and administration are explained, then the selection of verbs for valence analysis.
In general, in the anal-
ysis of the test responses, verb entries in Helbig and Schenkel's valence dictionary formed the basis for morpho-syntactic valence identification. To these identifications, our logico-semantic specifications were added, enabling us to form groups of semantically related verbal concepts and, in a word field-like approach, compare verbs within a group as well as different groups of verbs.
This approach enabled us to examine closely related verbal
concepts and how they are realized in different languages, native and target. At the same time, the limits and shortcomings of the logico-semantic valence specifications became known.
Some of the major areas treated are:
Agent-
Object-Experiencer realizations; Benefactive relations; Locative and Time relations; reflexivity and related problems, such as Stative passive versus state reflexive; adjectives and valence; lexicalization; focus; and 'function verb constructs'
(Funktionsverbgefuge).
In the last chapter, the psychological validity of relational constants is brought into focus, taking into consideration recent studies by Teigeler (1972), Engelkamp (1973), and Klntsch (1974). Based on these psychological insights, the study concludes with a proposal to introduce related verbs (predicates) in a 'cyclic' approach in which sentence-semantic concepts are gradually and systematically expanded.
155
ZUSAMMENFASSUNG
Diese Arbeit ist eine Untersuchung der linguistischen Valenz im morphosyntaktischen wie auch im logisch-semantischen Bereich. Die Anwendbarkeit des Valenzbegriffes in einer didaktischen Granmatik wird überprüft. Zuerst besprechen wir neuere linguistische Theorien und deutsche Granmatiken in bezug auf ihre Verwendung des Valenzbegriffes. Fehler, die in einem deutschen Sprachtest genacht wurden, unterstützen unsere Annahme, daß morphosyntaktische und logisch-semantische Valenz bei der Beschreibung semantischer Satzstrukturen eine wichtige Rolle spielen. Um Ähnlichkeiten wie auch Unterschiede zwischen semantischen Satzstrukturen innerhalb einer Sprache und van. Sprache zu Sprache—in unserem Falle Englisch und Deutsch—aufzuzeigen, wird die Verbindung norpho-syntaktischer und logisch-semantischer Valenz vorgeschlagen. Im ersten Abschnitt wird der Begriff 'pedagogical grammar' eingeführt, und Beziehungen zwischen einer didaktischen Grammatik und linguistischen Theorien werden erörtert. Um Vergleiche satzsemantischer Strukturen zwischen verschiedenen Sprachen zu ermöglichen, ist ein Beschreibungsverfahren notwendig, das universell, d.h. nicht nur einzelsprachlich, anwendbar ist. Das theoretische Beschreibungsmodell wird im zweiten Abschnitt erarbeitet. Ein Überblick stellt verschiedene Interpretationen des Valenzbegriffes vor. Dabei behandeln wir zuerst Tesnidre's Valenzbegriff und die Valenzkonzeption bei Erben, Brinkmann und Admoni. Anschließend werden Heibig und Schenkels Valenzwörterbuch, Engels Satzbaupläne und Vaters generatives Abhängigkeitsmodell besprochen. Nach einigen Bemerkungen zum Verhältnis zwischen Syntax und Semantik werden Kriterien festgelegt, die zur Abgrenzung morpho-syntaktischer Einheiten als Reflexionen semantischer Strukturen dienen. Insbesondere interessiert uns die Unterscheidung obligatorisch/fakultativ und frei in einer Valenztheorie sowie die Unterscheidung nuklear und peripheral in einer
156
Kasusgramnatik. Nachdem wir durch Berücksichtigung der Satzsemantik und der generativen Semantik Einblicke in die logisch-semantische Struktur der Sprache gewonnen haben, entwickeln wir 'Relationskonstanten'. Dabei werden die Relationskonstanten der 'Kausativgruppe1 mit Hilfe von Merkmalen identifiziert; bei der Bestimmung der übrigen Relationskonstanten verwenden wir die von Fillmore und Chafe erarbeiteten Modelle. Folgende Relationskonstanten werden verwendet: 'Cause1, 'Agent', 'Force1, 'Instrument', Object', 'Benefactive', 'Experiencer', 'Location', 'Time1, 'Final'. Diese Konstanten bilden die Basis für unser Modell einer Valenzanalyse im logisch-semantischen Bereich. Im dritten Abschnitt werden die im zweiten Abschnitt eingeführten Relationskonstanten bei der Analyse von Fehlern verwendet, die in einem deutschen Sprachtest gemacht wurden. Alle vier Teile des Tests enthielten Verbalausdrücke, die einem Lehrbuch der deutschen Sprache, 'Deutsch als Fremdsprache 1', entnommen sind. Zuerst werden Zielsetzung, Aufbau und Durchführung des experimentellen Sprachtests besprochen. Eine Liste der zur Analyse ausgewählten Verben (Prädikate) folgt. Die Eintragungen in Heibig und Schenkels Valen2wörterbuch bilden den Ausgangspunkt für die morpho-syntaktische Valenzbestinntung. Ein wortfeldähnlicher Angang ermöglicht es, verwandte Verbalbegriffe zu untersuchen und zu differenzieren. Gleichzeitig zeichnen sich dabei aber auch Grenzen einer logisch-semantischen Valenzbestimmung ab. Anschließend werden einzelne Verbgruppen und mit Valenz zusammenhängende Randgebiete ausführlich besprochen: 'Agent-Object-Experiencer Relation"; 'Beriefactive Relation'; 'Location'; 'Time1; Reflexivität und damit zusammenhängende Probleme wie Zustandspassiv und Zustandsreflexiv; Valenz bei Adjektiven; Lexikalisierung; Fokus; Funktionsverbgefüge. Im letzten Abschnitt wird die psychologische Validität von Relationskonstanten besprochen. Dabei verweisen wir auf Untersuchungen von Engelkamp, Kintsch und Teigeier. Unter Berücksichtigung dieser psychologischen Einblicke werden abschließend Möglichkeiten der Anwendung unserer Valenzuntersuchungen beim Erlernen von Satzstrukturen erörtert. Es wird z.B. ein zyklischer Angang vorgeschlagen, wobei verwandte Verbalbegriffe entsprechend ihrem Schwierigkeitsgrad systematisch eingeführt und erweitert werden.
157
APPENDIX A EXPERIMENTAL LANGUAGE TEST IN GERMAN
Part 1—Directions:
Example:
1.
2.
bleiben: 'to stay'
beginnen:
gehen:
Read each of the following communicative units in its entirety before you attempt to fill in blanks. Not every blank in the text (mini-dialog) requires a fill-in. Identify yourself with the situation in each unit. Underneath each verb write the English meaning it has in the particular context. A:
Warum
B:
Weil
A:
Haben die Vorlesungen schon
B:
Nein, die
A:
Wo
4.
haben:
hängen:
ich
du
nicht
kommen:
Sie? Hause.
A:
Fuß? Ja natürlich; es ist
nur zehn Minuten.
ihr Gepäck?
A: B:
Ja, wir ein Tasche.
A:
Ich
zwei Koffer
Mantel an d
B:
Wo
A:
An d
Tür.
A:
Wo
kommen Sie?
B:
Montag?
fangen erst nächste Woche
haben/sein?
5.
bis
da bereits wieder Vorlesungen
B:
B:
3.
bleibst
d
Frankfurt.
Mantel?
und
Tür
habe .
158
6.
wohnen:
A:
nicht mehr Ich ich habe jetzt selbst ein
Eltern,Wohnung
klein
Das ist aber eine Überraschung! 7.
i.
danken:
dauern:
A:
Ich möch
eure Hilfe danken.
B:
Gern geschehen.
A:
Wie lange _____
B:
Noch ]
9.
fragen:
A:
W
geben:
A:
Polizisten. W
hast du das Telegramm Lehrerin
B:
11.
12.
gehören:
haben:
A:
Gehört d
B:
Ja,
A:
verkaufen:
Kamera Stefan?
ich glaube,
^^__ (Pronomen)
gehört (Pronomen)
Sie den Unfall?
Wo
Autobahn
B:
13.
2 Stunde
fragt Ernst?
B:
1O.
der Bus kommt?
München und Stuttgart.
A:
Sollen wir das Haus nicht später verkaufen?
B:
Nein, es
___ jetzt (Passiv)
müssen 14.
15.
kommen:
lesen:
B:
Entschuldigen Sie, wann Bus Bonn? 30 _ l halb
A:
Hast du Erikas Brief schon
B:
Nein, d
der nächste
erst später
(Futur) 16.
liegen:
A:
ich
lange müssen
B:
Mind
eine Woche!
159
17.
18.
anprobieren:
schenken:
Sommerkleid
gelb zu groß
A:
Ich habe d_ leider ist
B:
Sie doch mal d Dann Hose und d schwarz_ passen die.
A:
Wer hat d_
B:
Die
rot
die Blumen m
von Peter
sein/haben? 19.
2O.
sein:
sprechen:
(Passiv)
A:
Ist das schon wieder ein
B:
Aber nein, die Großmutter!
neu
Bluse?
schon sehr alt,
von mein
Inge gut Deutsch?
A: B:
Pulli , vielleicht
Sicher, sie
von von uns
(Superlativ) all
21.
werden:
A:
Über Nacht
es kalt sein/haben?
Da hast du recht! 22.
23.
zeigen:
fahren:
Sie m_ _
A: B:
Welch
A:
Wann f
B:
Vor einig
?
D
bitte d
Kleid dort.
grün_ ? hier ein Bus Minute
Bonn ist gerade ein
24.
stellen:
A:
Ich kann hier keinen Parkplatz finden; darf ich mein Wagen vor ihrem Haus
2 5.
kommen:
A:
Wieviel Wagen?
B:
Er hatte Glück, er hat noch 6OO DM
A:
ist der Fuß
B:
Ich glaube nicht, aber wir wollen lieber röntgen.
26.
27.
brechen:
aussteigen:
A:
sein
Sie in Köln sein/haben?
B:
Nein, in Frankfurt.
alt
doch
160
28.
29.
essen:
fliegen:
A:
er gestern in der "Sonne"?
B:
Er
A:
Onkel Otto kam gestern aus den USA an.
B:
Bleibt er einige Zeit?
doch immer da!
Nein, er Wien 3O.
gefallen:
A:
schon heute früh Ihr
Wie gef
Eltern das neue Haus? sehr!
gef
B:
(Pronomen)
(Pronomen) 31.
grüßen:
A:
Kamen all
B:
Ja, sie w
eingeladen
Gast
?
von Herrn Schneider (Passiv)
32.
einkaufen:
A:
Hier
B:
Wirklich? b
billig
man
Ich habe im "Kaufhof" noch (Komparativ)
32.
kommen:
A:
Jochen, wo bist du? Keller!
B:
doch bitte 34.
35.
kosten:
liegen/legen:
A:
Alles ist teuer gew
B:
Ja, früher 6O DM
A: Wo Wohnzimmer?
so eine Uhr nur wir
Teppich,
Nein, da 36.
reparieren:
auf!
schon ein
A:
Ich möch
mein Auto reparieren lassen.
B:
Ich glaube, das kann nicht mehr (Passiv)
37.
38.
setzen/sitzen: A:
unterhalten:
doch in
B:
Ich
A:
Wo hab unterhalten?
B: A:
lieber ihr
Garten! der Terrasse. ganz
Politik. Da
spreche ich lieber nicht.
Abend
161
39.
rufen:
B:
4O.
41.
ziehen:
arbeiten:
ruf ra
A:
bitten morgen
Ist es früh genug, wenn Uhr
acht ?
A:
Ist d
nicht zu warm?
B:
Aber nein, sonst hätte ich meine Jacke
A:
Arbeiten Sie auch samstags?
B:
Ja, leider
bei uns auch samstags (Passiv)
42.
43.
bleiben:
einladen:
A:
Vielleicht finden Sie ein Zimmer; bleiben Sie dann?
B:
Ja, wenn_ ich
A:
Ihr wollt sicher auch Prof. Klug einladen!
B:
Wir denken nicht
Zimmer
, Prof. Klug
(Infinitiv) 44.
empfangen:
A:
Was hatte Erika zu erzählen?
B:
Sie erzählte, daß sie
Direktor (Passiv)
45.
Er freut Geschenk.
freuen: B:
46.
haben:
sicher ü
Hoffentlich, da
A:
groß
würde ich
du kein
auch freuen.
lang
Leiter? (Komparativ)
Leider nicht, das , d~ (Superlativ) 47.
gelegen/ gelegt:
Ich
die Brille doch auf , wo ist die denn?
Hier, sie 48.
nehmen:
A:
Welch
B:
Ich nehme d
ist d ich habe.
Stuhl
auf
Rock
du?
kurz (Komparativ)
Tisch
162
49.
50.
sitzen/setzen
stehen/ stellen:
A: Wo
sollen wir Fenster, wo wir eine schöne
B:
Am liebsten Aussicht haben.
A:
Soll ich das Fahrrad in
B:
Ja, da
__ Garage
auch die ander
Räder.
Part 2—Directions: Use all words (phrases) given with a particular verb plus any additional material necessary to form a German sentence. You decide word order, endings, etc. A certain adverb might require the use of a particular tense (present, past or f u t u r e ) ! Example:
sein:
Zeitung - unter Tasche - hier Die Zeitung ist hier unter der Tasche.
or
Hier ...
1.
warten:
Mutter - Stunde - Arzt - schon
2.
schenken:
schön Buch - kürt - letzte Woche - Freundin
3.
gehen:
Daniel - neulich - Bahnhof - müssen
4.
begleiten:
Sängerin - Pianist - gestern abend (Passiv)
5.
besuchen:
Jochen - vorher - vier Jahre - Volksschule
6.
abholen:
Taxi - Direktor - vorgestern - Flughafen
7.
anmelden:
Verkäufer - Büro - schon
8.
bleiben:
Wetter - ganz Tag - gestern - schön
9.
bringen:
Eltern - Kinder - letzte Woche - aufs Land
l.O.
helfen:
Peter - Professor - schwer Bücher - tragen
11.
erwarten:
Kellner - vorige Woche - Chef - Brief
12.
vergessen:
Ute - Schlüssel - Büro - neulich
13.
ansehen:
Student - einige Zimmer - morgen
14.
geben:
Post - Sekretärin - Direktor - gestern - nicht
15.
stattfinden:
vergangenen Sonntag - Stadthalle - Chorkonzert
(Part 3 of the experimental test is not included—see 3.1.4).
163
Part 4—Directions; This part is not to test your "translation skill"! The German equivalent may d i f f e r considerably in structure. Focus on the verb! It is important that you. attempt to do all items! Example:
What is that supposed
to mean?
'Was soll das heißen?'
1.
Please, answer me!
2.
I'm getting hungry1.
3.
She broke her arm.
4.
Has he not taken home?
5.
Please, remember that there is no lecture tomorrow.
6.
Have you ever flown?
7.
They were questioned by the police.
8.
I'm pleased that you could visit us.
9.
He was given a new car.
10.
Give my regards to your parents.
11.
He would like to have a Porsche.
12.
If
13.
Mho did he buy that old car from?
14.
Do you know Vienna?
15.
It was a pleasure to make your acquaintance.
16.
Wouldn't that be nice?
17.
I rather take the table by the window.
18.
Please, take a seat!
19.
They met in front of the library.
20.
He was sorry he could not swim.
21.
Who paid the bill?
22.
Show me your passport, please!
23.
Does the weather change often
24.
Karl watched the entire soccer game.
25.
He likes to eat fish but it does not agree with him.
only we could help him!
here?
164
26.
Thank you for the kind letter.
27.
Did he apologize?
28.
What's wrong with him?
29.
How did you like the trip?
30.
Whose key is this?
31.
I can't believe him.
32.
Are you in a hurry?
33.
She asked for their address.
34.
He has changed.
35.
Have you answered his letter yet?
36.
I'm getting tired of him.
37.
Do you know who broke the vase?
38.
I don't mind.
39.
How did the cake taste?
40.
Years ago he climbed many mountains.
41.
Fortunately, nobody was injured.
42.
I really did not expect you so early.
43.
She asked me a question.
44.
There was no parking place.
45.
May I be excused?
(To whom does this key belong?)
165
APPENDIX B ALPHABETICAL LIST OF VERBS (PREDICATES) ANALYZED AND DISCUSSED IN CHAPTER 3
Test Item
Discussed on Page
abfahren
depart
1.23
103,109,110
abfliegen
depart (by plane)
1.29
103
abholen
pick up
2.6
103,109
abkaufen
buy from
abstellen
park, leave
ähneln
resemble
134
ähnlich sein
be similar to
134
ändern
change
(ver-) sich
101
1.24
4.23, 4.34
108
126
ändern
change
126
angehören
belong to, be a member of
87,89
ärgern
annoy
116,117
ärgern sich
be annoyed, angry
116,117
anmelden (sich)
report, register
anschaffen sich
set for oneself, obtain
ansehen
(sich)
2.7
127
100,101
look at (oneself)
2.12
122,123,125
ansehen- sich
watch, look at
4.24
123,125
antworten
answer
4.1
68-70
anziehen
put on (clothes)
122
aufsetzen
put on (cap, hat)
122
aussteigen
get off (train, car)
1.27
103,110,111
ausziehen
take o f f , undress
1.4O
121,122
166
Test Item
Discussed on Page
4,35
68-70
beantworten
answer
beeilen sich
hurry (up)
113
bedanken sich
thank for
92
befinden sich
be situated, located
106,121
befragen
question, interrogate
68,72,73
befragen sich
inquire
begeben sich
go (to)
113
begegnen
meet (by chance)
130
begießen
water
136
begrüßen
greet, welcome
bekleben
glue, paste
136
bekommen.
get (receive)
89,91
bekommen
get (buy)
90,91
bekommen.
get
90,91,130
bekommen.
agree with (suit the health of)
92
beobachten
watch, observe
124,125
bepflanzen
plant
136
berauben
rob
102
obtain, get
101
beschmieren
smear
135,136
besitzen
own
87,89
besichtigen
visit, look at
125
bestehlen
steal from
102
besteigen
climb (to the top)
bestreuen
sprinkle
135,136
betrachten
watch, observe (carefully)
124,125
bewegen (sich)
move
142
bleiben
stay, remain
beschaffen
(sich)
4.7
1.31
(become)
4.4O
2,8
73
84
103,134
143
bleiben in Kraft
stay in effect
144
in Mode
stay in fashion
142
bleich sein
be pale
139
brechen
break
1,26, 4.3
137,139
167
bringen
l bringen.
take to
Test Item
Discussed on Page
4.4
95,96
bring
96
in Bewegung
set in motion
141
in Mode
bring into fashion
142
in Verlegenheit
embarrass
145
in Wut
enrage
143
zum Schluß
bring to a conclusion
141
zur Aufführung
perform
141
bringen
danken
thank
1.7
92
dauern
last
1.8
130
denken sich
think (oneself)
denken an
remember
einschlafen
fall asleep
empfangen,
welcome (receive)
84,86
empfangen-
receive (take)
86
entschuldigen (sich)
excuse (oneself)
127
erfragen
find out
73
erfrieren
freeze to death
140
erinnern, sich
remember
erinnern-
remind
117,127
erkundigen sich
inquire
73
ermüden
get tired
139
erröten
blush
139
erschießen
shoot to death
139
erschlagen
beat to death
139
ertrinken
drown
139
erwachen
awaken, wake up
139
erzählen
tell
75,82,83
fahren
drive, travel
110
fliegen
fly
103,109
fragen
ask
1.9
68,72,73
Frage (eine Frage stellen)
ask a question
4.43
68,72
128 4.5
4.5
127
117,127
168
Test Item
Discussed on Page
frieren
freeze
140
fühlen sich
feel
131
fürchten
fear
1 1 4 , 1 20
fürchten sich
be afraid
114,119
geben
give
1.10, 2.14 4.9
93,94
gehen
go
1.2, 2.3
103
gehören
bei ing to
1.11, 4.30
86,89
gehören zu
be part of
88,89
belong (have its proper place)
88,89
fly into a rage
143
gießen
water
136
grüßen
greet
grüßen lassen
send regards
84
gut sein
be good to
133
hängen
hang (be hanging)
1O6, 1O8
hängen-
hang
1O3, 1O6, 107,108
helfen
help
93
holen
get
Hunger bekommen
get hungry
hungrig werden
get hungry
interessiert sein
be interested in
133
kalt sein
be cold
131
kaufen
buy
kleben
glue
kommen
come
gehören i n ( a u f ,
..
geraten in Wut
4.10
84
(pick up)
109 131 4.2
131
4.13
97,100, 101,117 136
1.5,
1.33
95
kommen in Bewegung
start moving
142
in Mode
come into fashion
142
in Wut
get enraged
143
zur Aufführung
be performed
142
zum Schluß
come to an end
142,144
zur Sprache
come up for debate
144
169 Test Item zum Vortrag
be recited
Discussed on Page 144
legen
put, lay
lieb sein
be dear to
liegen
lie,
müde sein
be tired
müde werden
get tired
nehmen.l nehmen»
take
96
take to
97
in Anspruch
put forward a claim to
145
zur Kenntnis
take cognizance of
145
pflanzen
plant
135,136
Platz nehmen
take a seat
rauben
rob
102
reden
speak, talk
75,78,83
reparieren
repair
118
rot sein
be red
139
rot werden
turn red, blush
139
sagen
say
75,78,79, 80,83
schämen sich
be ashamed
113
schenken
give a present
schlagen
beat
139
schießen
shoot
139
schmieren
smear
135,136
schreiben
write
118
schwärmen
swarm
135
sehen
see
123,125
in Kraft
be in effect
144
in Mode
be in fashion
142
in Wut
be enraged
143
be located
1.35
103,107,108 133
1.35
103,104,108 132,139
4.36
132
nehmen
4.18
2.2
103
95
sein
setzen
put,
seat
1.37, 1.49
103,107,108
setzen in Kraft
put in effect
144
außer Kraft
annul
144
170
Test Item
Discussed on Page
sitzen
sit
1.37, 1.49
103,105,108
sprechen
speak
1 . 2O
75-77,83
stattfinden
take place
2.15
130
stehen
stand
1 . 5O
1O3,1O4,1O8
stehen unter Aufsicht
be under supervision
143
stehlen
steal
102,110
steigen
climb, ascent
134
stellen
put, position
1 . 5O
1O3,1O7,1O8
stellen unter Aufsicht streuen treffen,
(sich)
treffen-
put under supervision
143
sprinkle
135,136
meet (each other)
4.19
103,129,130
meet (by chance)
129
become effective
144
trinken
drink
139
verändern (sich)
change
verkaufen
sell
98,99,101, 117,120
verkaufen
sell (carry)
98
treten in Kraft
verletzen
(sich)
injure
4 . 23 , 4 . 34
4.41
126,127
128
verletzen^
hurt (feelings)
129
verlieben sich
fall in love
118,120
verliebt sein
be in love
118/120
vorstellen
imagine
115,128
vorstellen,, (sich)
introduce (a person)
115
wach sein
be awake
139
waschen
wash
112,114,118, 119,129
wegnehmen
take away
98
wimmeln
teem with
135
wohnen
live, reside
103
zerbeißen
crunch, bite through
137
zerbrechen
break to pieces
sich
4.37
137,139
171
Test Item
Discussed on Page
zerschneiden
cut to pieces
137
zerdrücken
crush
137
zerfressen
gnaw
137
zerlegen
dissect, decompose
137
zerreißen
tear to pieces
137
zerstören
destroy
137
zufrieden sein
be satisfied
131 132
zulegen sich
get for oneself
loi
zusehen
observe, watch
124 125
172
BIBLIOGRAPHY
Abraham, Werner. 1971. Kasustheorie. Frankfurt am Main: Athenäum Verlag GmbH. Admoni, Wladimir G. 197O. Der deutsche Sprachbau. 3rd ed. München: C. H. Beck'sehe Verlagsbuchhandlung. . 1972. 'Grammatiktheorien in ihrer Anwendung auf das heutige Deutsch 1 . Sprache der Gegenwart, vol. 2O. Edited by Hugo Moser. Düsseldorf: Pädagogischer Verlag Schwann. . 1974. 'Die Satzmodelle und die logisch-grammatischen Typen des Satzes'. Deutsch als Fremdsprache, l, 34-42. Anderson, John M. 1971a. 'Dependency and Grammatical Function". Foundations of Language, l, 3O-73. . 1971b. The Grammar of Case: Towards a Localistic Theory. Cambridge: Cambridge University Press. Ballweg, Joachim, Hacker, Hans Jürgen, and Schumacher, Helmut. 1971. 'Semantik und Satzbaupläne 1 . Muttersprache, 81, 224-34. Baumgärtner, Klaus. 197O. 'Konstituenz und Dependenz 1 . Wege der Forschung, vol. 46. Edited by Hugo Steger. Darmstadt: Wissenschaftliche Buchgesellschaft. Behaghel, Otto. 1924. Deutsche Syntax, vol.
2. Heidelberg: C. Winter.
Betterridge, Harold T. 1962. The New Cassell's German Dictionary. New York: Funk & Wagnalls Company. Bierwisch, Manfred. 1965. Grammatik des deutschen Verbs. 2nd ed. Berlin: Akademie-Verlag. Braun, Korbinian; Nieder, Lorenz; and Schmöe, Friedrich. 1967. Deutsch als Fremdsprache 1. Stuttgart: Ernst Klett Verlag. Brekle, Herbert E. 197O. Generative Satzsemantik und transformationelle Syntax im System der englischen Nominalkomposition. München: Wilhelm Fink Verlag. Brinker, Klaus. 1972. Konstituentenstrukturgrammatik und operationale Satzgliedanalyse: methodenkritische Untersuchungen zur Syntax des einfachen Satzes im Deutschen. Frankfurt am Main: Athenäum Verlag GmbH. Brinkmann, Hennig. 1971. Die deutsche Sprache - Gestalt und Leistung. 2nd ed. Düsseldorf: Pädagogischer Verlag Schwann. Bühler, Karl. 1934. Sprachtheorie - Die Darstellungsfunktion der Sprache. Jena: Verlag von G. Fischer.
173 Buscha, Joachim. 1972. 'Zur Wortklassenbestimmung der Reflexiva in der deutschen Gegenwartsprache 1 . Deutsch als Fremdsprache, 3, 151-59. Chafe, Wallace. 1970. Meaning and the Structure of Language. Chicago: University of Chicago Press. . 1971a.
'Directionality and Paraphrase 1 . Language, 47,
1-26.
1971b. "Linguistics and Human Knowledge'. Monograph Series on Languages and Linguistics. Number 24. Edited by Richard J. O ' B r i e n . Washington, D. C.: Georgetown University Press. Chomsky, Noam. 1965. Aspects of the Theory of Syntax. Cambridge, Massachusetts: The M . I . T . Press. . 1972. 'Deep Structure, Surface Structure and Semantic Interpretation'. Studies on Semantics in Generative Grammar. Series Minor 1O7 of Jariua Linguarum. The Hague: Mouton f* Co. Cook, Walter A. 1971. 'Improvements in Case Grammar 1 . Languages and Linguistics: Working Papers. Number 2. Washington, D. C . : Georgetown University School of Languages and Linguistics. . 1972. Set of Postulates for Case Grammar Analysis'. Languages and Linguistics: Working Papers. Number 4. Washington, D. C - : Georgetown University School of Languages and Linguistics. . 1973. 'Covert Case Roles'. Languages and Linguistics: Working Papers. Number 7. Washington, D. C . : Georgetown University School of Languages and Linguistics. Corder, Pit. 1974. 'Problems and Solutions in Applied Linguistics'. Proceedings of the Third Congress of the Association Internationale de Linguistique Appliquee (Copenhagen, 1 9 7 2 ) , vol. 3. Heidelberg: Julius Groos Verlag. DiPietro, Robert J. 1971. Language Structures in Contrast. Rowley, Massachusetts: Newbury House Publishers. Engel, Ulrich. 197 . 'Die deutschen Satzbaupläne'. Wirkendes Wort, 6, 361-92. Engelen, Bernhard. 1968. 'Zum System der Funktionverbgefüge'. Wirkendes Wort, 5, 289-3O3. Engelkamp, Johannes. 1973. Semantische Struktur und die Verarbeitung von Sätzen. Bern: Verlag Hans Huber. Erben, Johannes. 1972. Deutsche Grammatik - Ein Abriß, München: Max Hueber Verlag.
l l t h ed.
Fillmore, Charles. 1966. Proposal Concerning English Prepositions'. Monograph Series on Languages and Linguistics. Number 19. Edited by Francis P. Dinneen. Washington, D. C . : Georgetown University Press. . 1968. 'The Case for Case'. Universals in Linguistic Theory. Edited by Emmon Bach and Robert T. Harms. New York: Holt, Rinehard and Winston.
174
. 1971. 'Some Problems for Case Grammar 1 . Monograph Series on Languages and Linguistics. Number 24. Edited by Richard J. O'Brien. Washington, D. C . : Georgetown University Press. . 1974. 'The Future of Semantics 1 . Lecture at the LSA Golden Anniversary Symposium. Amherst, Massachusetts. (Personal tape recording). Fink, Stefan R. 1976. 'Semantic-Pragmatic Aspects in Foreign Language Pedagogy Based on Case Grammar and Valence Theory'. Linguistische Berichte, 41, 77-87. . 1977 (in press). 'Analytical Trends in Contemporary German: A Semantic Description of "Function Verbs" expressing Aktionsarten "Modes of Events" 1 . The Third Lacus Forum 1976. Columbus, S . C . : Hornbeam Press, Inc. Finke, Peter. 1974. Theoretische Probleme der Kasusgrammatik. Kronberg: Scriptor Verlag. Fourquet, Jean. 1976. Zur Sonderstellung des Subjekts. Eine Auseinandersetzung mit Tesniere und Chomsky. Wirkendes Wort, 26, 234-40. Fries, Charles C. 1952. The Structure of English. New York: Barcourt, Brace & Co. Glinz, Hans. 1962. Die innere Form des Deutschen - eine neue deutsche Grammatik. 3rd ed. Bern: A. Francke Verlag. Götze, Lutz. 1973. Review of Wörterbuch zur Valenz und Distribution deutscher Verben, by G. Heibig and W. Schenkel. Zielsprache Deutsch, 3, 137-38. . 1974. 'Zu den Begriffspaaren "Obligatorisch/fakultativ" und "notwendig/nicht notwendig" in einer Valenzgrammatik und ihrer Relevanz für den Sprachunterricht 1 . Zielsprache Deutsch, 2, 62-71. Grebe, Paul. 1966. Duden - Grammatik der deutschen Sprache. 2nd ed. Mannheim: Bibliographisches Institut, Duden Verlag. Günther, Hartmut. 1974. Das System der Verben mit BE- in der deutschen Sprache der Gegenwart. Ein Beitrag zur Struktur des Lexikons der deutschen Sprache. Tübingen: Max Niemeyer Verlag. Halliday, M. A. K. 1971. 'Language Structure and Language Function 1 . New Horizons in Linguistics. Edited by John Lyons. Harmondsworth, England: Penguin Books, Ltd. Harris, Zellig S. 1957. 'Co-occurrences and Transformation Structure*. Language, 33, 283-34O.
in Linguistic
Hays, David G. 1964. 'Dependency Theory: A Formalism and Some Observations'. Language, 4O, 511-25. Heger, Klaus. 1966. 'Valenz Diathese und Kasus 1 . Zeitschrift für Romanische Philologie, 82, 1/2, 138-7O. . 1971. Monem, Wort und Satz. Tübingen: Max Niemeyer Verlag.
175
Heibig, Gerhard. 1971a. "Untersuchungen zur Valenz und Distribution deutscher Verben 1 . Probleme der Sprachwissenschaft. Series Minor 118 of Janua Linguarum. The Hague: Mouton & Co. . 1971b. 'Zu einigen Spezialproblemen der Valenztheorie 1 . Deutsch als Fremdsprache, 5, 269-82. . ed. 1971c. Beiträge zur Valenztheorie. Series Minor 115 of Janua Linguarum. The Hague: Mouton & Co. . 1976. "Valenz Semantik und Satzmodelle". Deutsch als Fremdsprache, 2, 99-1O6. and Buscha, Joachim. 1972. Deutsche Grammatik. Ein Handbuch für den Ausländerunterricht. Leipzig: VEB Verlag Enzykopiidie. and Kempter, Fricz. 1973a. Das Zustandspassiv. Leipzig: VEB Verlag. and Schenkel, Wolfgang. 1973b. Wörterbuch zur Valenz und Distribution deutscher Verben. 2nd ed. Leipzig: VEB Verlag. Heringer, Hans-Jürgen. 1967. "Wertigkeit und nullwertige Verben im Deutschen". Zeitschrift für Deutsche Sprache, 23, 1/2, 14-34. . 197Oa. Theorie der deutschen Syntax. München: Max Hueber Verlag. . 197Ob. Deutsche Syntax. Berlin: De Gruyter. and Öhlschläger, Günther. 1973. 'Wertigkeitstheorie und Sprachunterricht". Angewandte Sprachwissenschaft und Deutschuntericht. Edited by Gerhard Nickel. München: Max Hueber Verlag. Herrlitz, Wolfgang. 1973. Funktionsverbgefüge vom Typ "in Erfahrung bringen"; ein Beitrag zur generativ-transformationellen Grammatik des Deutschen. Tübingen: Max Niemeyer Verlag. Heyse, Johann C. A. 19O8. Deutsche Grammatik. Hannover: Hahn Verlag. Hjelmslev, Louis. 1961. Prolegomena to a Theory of Language. Revised English Edition. Madison: University of Wisconsin Press. Hockett, Charles F. 1958. A Course in Modern Linguistics. Toronto.· The Macmillan Company. Huddleston, Rodney. 197O. "Some Remarks on Case Grammar". Linguistic Inquiry, l, 5O1-11. Hüllen, Werner. 1972. 'Zur linguistischen Begründung fremdsprachlicher Übungsfonnen". Linguistik und Didaktik, 9, 32-41. Jacobs, Roderick A. and Rosenbaum, Peter S. 1968. English Transformational Grammar. Waltham, Massachusetts: Blaisdell Publishing Company. Jäger, Gert. 1974. "Zur Diskussion über das Verhältnis von Sprachwissenschaft und Fremdsprachenunterricht aus linguistischer Sicht". Deutsch als Fremdsprache, l, 25-33.
176
Jakobovits, Leon A. 197O. Foreign Language Learning: A Psycholinguistic Analysis. Rowley, Massachusetts: Newbury House Publishers. Joos, Martin. 1964. The English Verb. Madison: University of Wisconsin Press. Jung, Walter. 1967. Grammatik der deutschen Sprache. Leipzig: VEB Bibliographisches Institut. Kintsch, Walter. 1974. The Representation of Meaning in Memory. New York: John Wiley & Sons. König, Ekkehard. 1972. 'Anmerkungen zum Problem der Beschreibung von R e f l e x i v i t ä t ' . Linguistische Berichte, 19, 33-37. Kolvenbach, Monika. 1972. 'Verbvalenzuntersuchungen: Eine Voraussetzung für die Monosemierung von Verbinhalten'. Linguistische Studien, 2 (Sprache der Gegenwart, vol. 2 2 ) . Edited by Hugo Moser. Düsseldorf: Pädagogischer Verlag Schwann. Lado, Robert. 1961. Language Testing: The Construction and Use of Foreign Language Tests. London: Longmans, Green & Co. . 1964. Language Teaching: A Scientific Approach. New York: McGraw-Hill, Inc. . 197O. 'Language, Thought and Memory in Language Teaching: A Thought V i e w ' . The Modern Language Journal, 8, 580-85. Leech, Geoffrey N . 197O. Towards a Semantic Description of English. Bloomington: Indiana University Press. . 1974. Semantics. Harmondsworth, England: Penguin Books, Ltd. Leisi, Ernst. 1967. Der Wortinhalt - Seine Struktur im Deutschen und Englischen. 3rd ed. Heidelberg: Quelle & Meyer. . 1973. Praxis der englischen Semantik. Heidelberg: C. Winter Universitci tsverlag. Lyons, John. 1968. Introduction to Theoretical Linguistics. Cambridge: Cambridge University Press. Mattutat, Heinrich. 1969. Deutsche Grundsprache - Wort- und Satzlexikon. Stuttgart: Ernst Klett Verlag. Meiner, Johann W. 1781. Versuch einer an der menschlichen Sprache abgebildeten Vernunftslehre,· oder. Philosophische und allgemeine Sprachlehre. Leipzig. Mellema, Paul. 1974. Brief Against Case Grammar'. Foundations of Language, 11, 39-76. Moser, Hugo, ed. 197O. Probleme der kontrastiven Grammatik. Düsseldorf: Pädagogischer Verlag Schwann. , ed. 1972. Neue Grammatiktheorien und ihre Anwendung auf das heutige Deutsch. Düsseldorf: Pädagogischer Verlag Schwann.
177
Nickel, Gerhard, ed. 1972. Reader zur kontrastiven L i n g u i s t i k . F r a n k f u r t am Main: Antheüum Verlag GmbH. Nilsen, Don L. F. 1972. Toward a Semantic Specification of Deep Case. Series Minor 152 of Janua Linguarum. The Hague: Mouton & Co. . 1973. The Instrumental Case in English. Series Minor 156 of Janua Linguarum. The Hague: Mouton & Co. Oiler, John W. 1974. 'Expectancy for Successive Elements: Key Ingredients to Language U s e 1 . Foreign Language Annals, 4, 443-51. Perl, Matthias. 1973. 'Einige Gedanken zur semantischen Valenz des deutschen Verbs " v e r k a u f e n " ' . Deutsch als Fremdsprache, 2, 102-07. . 1974. 'Wissenschaftliche Arbeitstagung: Valenz und Ä q u i v e l e n z ' . Deutsch als Fremdsprache, 4 , 241-43. Platt, John T. 1971. Grammatical Form and Grammatical Meaning—A Tagmemic View of Fillmore's Deep Structure Case Concepts. Amsterdam: North-Holland Publishing Company. Pusch, Luise F. 1972. 'Bemerkungen über partitive und holistische Konstruktionen im E>eutschen und E n g l i s c h e n ' . Reader zur kontrastiven L i n g u i s t i k . Edited by Gerhard N i c k e l . F r a n k f u r t am Main: Athenäum Verlag GmbH. Robinson, Jane J. 1969. 'Case, Category, and C o n f i g u r a t i o n 1 . Journal of Linguistics, 6, 57-80. Schmidt, Wilhelm. 1963. Lexikalische und aktuelle Bedeutung; ein Beitrag zur Theorie; der Wortbedeutung. Berlin: AkademieVerlag. Schulz, Dora and Griesbach, Heinz. 1970. Grammatik der deutschen Sprache. 8th ed. München: Max Hueber Verlag. Schumacher, Helmut. 1972. 'Zum deutschen V a l e n z l e x i k o n 1 . Sprache der Gegenwart, vol. 2O. Edited by Hugo Moser. Düsseldorf: Pädagogischer Verlag Schwann. , ed. 1975. Untersuchungen zur Verbvalenz. Tübingen: Narr. Seiler, Hansjakob. 1973. Linguistic Workshop l - Vorarbeiten zu einem Universalienprojekt. München: Wilhelm Fink Verlag. Seuren, Pieter A. M. 1969. Operators and Nucleus—A Contribution to the Theory of Grammar. Cambridge: Cambridge University Press. . 1973. 'Autonome und semantische S y n t a x ' . Generative Semantik Semantische Syntax. Edited by Pieter A. M . Seuren. Düsseldorf: Pädagogischer Verlag Schwann.
178 Sommerfeldt, Karl Ernst. 1971. 'Zur Valenz des Adjektivs 1 . Deutsch als Fremdsprache, 2, 113-17. and Schreiber, Herbert. 1974. Wörterbuch zur Valenz und Distribution deutscher Adjektive. Leipzig: VEB Bibliographisches Institut. Starosta, Stanley. 1973. 'The Faces of Case'. Language Sciences. Number 25. Bloomington: Indiana University Research Center for the Language Sciences. Stötzel, Georg. 197O. Ausdrucksseite und Inhaltsseite der Sprache Methodenkritische Studie am Beispiel der deutschen Reflexivverben. München: Max Hueber Verlag. Strauch, R. O. U. 1972. "Bemerkungen zum Verhältnis von theoretischer Linguistik und Fremdsprachendidaktik 1 . Linguistik und Didaktik, 9, 2O-31. Teigeier, Peter. 1972. Satzstruktur und Lernverhalten. Bern: Verlag Hans Huber. Tesniere, Lucien. 1953. Esquisse d ' u n e syntaxe structurale. Paris: C. Klincksieck. . 1959. Elements de syntaxe structurale. Paris: C. Klincksieck. Vater, Heinz. 1973. Dänische Subjekt-und Objektsätze: Ein Beitrag zur generativen Dependenzgrammatik. Tübingen: Max Niemeyer Verlag. . 1975a. 'Towards a Generative Dependency Grammar 1 . Lingua, 36, 121-45. . 1975b. 'Zur Verbvalenz 1 . Kwartalnik Neofilologiczny, 22/3, 291-306. Zarechnak, Michael. 1974. 'Reflections on Cook's Case Grammar Matrix 1 . Working Papers on Languages and Linguistics. Number 8. Washington, D. C.: Georgetown University School of Languages and Linguistics. Zifonum, Gisela. 1972. 'Über die Unverträglichkeit verschiedener Valenzbegriffe und ihre Verwertbarkeit in semantischen Beschreibungen 1 . Zeitschrift für Dialektologie und Linguistik, 2, 171205. . 1973. Zur Theorie der Wortbildung am Beispiel deutscher Präfixverben. München: Max Hueber Verlag. Zimmerman, Rüdiger. 1972a. "Subjektlose und intransitive Sätze im Deutschen und ihre englische Äquivalente". International Review of Applied Linguistics, 3, 233-45. . 1972b. "Die Kasusgrammatik in der Angewandten und Kontrastiven Linguistik". International Review of Applied Linguistics, 2, 167-78.
179
ABBREVIATIONS AND SYMBOLS
A
Agent
a
accusative case
aff
affective
B
Benefactive
C
Cause
d
dative case
DIR
Direction
(Source)
Direction
(Goal)
0
DIR
(j
E
Experiencer
eff
effective
EV
close verbal complement (enge Verbergänzung)
F
Force
FN
Final (dativus commodi)
FV
free verbal complement (freie Verbergänzung)
G
Goal
g
genitive case
HuS
Helbig and Schenkel's valence dictionary
I
Instrument
Inf
Infinitive with 'to'
L
Location
L
Location (Direction)
L .
Location (Direction-Source)
L
Location (Direction-Goal)
n
nominative case
NS, „ , aaB, ot>, w
clause (introduced by that, i f , WH)
UXlTo
180
ο
Object (Relation)
Ob
Object (Grammatical)
Ο
Object
Id
(Identification)
Object (inalienable possession)
Inal
Object (Total inclusion)
Incl O Meas
Object (Measure) Object (Partial inclusion)
°Part
Object (Resemble)
°Res P
preposition
pS
prepositional phrase
pSa
prepositional phrase (accusative case)
Refl
Reflexive
S
Source
Sa
noun phrase (accusative case)
Sn
noun phrase (nominative case)
S,
noun phrase (dative case) Time Time (Duration of time)
Dur
Time (Point in time)
Punct
Verb (Variant 1) A, 1,0
deep case frame marker
•p
Not generally accepted sentence (phrase)
*
Non-acceptable sentence
(phrase)
alternate possibility facultative occurrence l + (1) = 2
number of obligatory and facultative positions (Mitspieler) and quantitative valence of verb ( 2 ) . (This specification appears as subscript after the verb) becomes
l
l
facultative reflexive
l
l
obligatory reflexive