Aristotle on Sleep and Dreams: A Text and Translation with Introduction, Notes, and Glossary [reprint ed.] 0856686743, 9780856686740, 9780856686757

This work is designed to make Aristotle's neglected but fascinating writings on sleep and dreams accessible in tran

137 97 12MB

English Pages 202 [223] Year 1996

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover
Title Page
Copyright Page
Contents
Acknowledgments
Preface
Note on the Text and Translation
Abbreviations
Introduction
1. Aristotle's Life and Work
2. The Parva Naturalia
3. The Greek Concept of Dreaming
4. Plato's Legacy
5. The De Somno and the De Insomniis
6. Dreams and the Imagination
7. Dreaming and Teleology
8. The Function of Dreams
9. Aristotle and Freud
10. Dreams and the 'Daemonic'
11. Aristotle and Malcolm
12. Conclusion
Text and Translation
On Sleep and Waking
On Dreams
On Divination through Sleep
Notes
On Sleep and Waking
On Dreams
On Divination through Sleep
Appendix: Aristotle's Historia Animalium, IV. 10
Glossary
Seiect Bibliography
Index
Recommend Papers

Aristotle on Sleep and Dreams: A Text and Translation with Introduction, Notes, and Glossary [reprint ed.]
 0856686743, 9780856686740, 9780856686757

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

ARIS & PHILLIPS CLASSICAL TEXTS

ARISTOTLE On Sleep and Dreams

EDITED WITH AN INTRODUCTION, TRANSLATION & COMMENTARY

David Gallop

BY

Aris and Phillips Classical Texts

ARISTOTLE

On Sleep and Dreams

David Gallop

Aris & Phillips is an imprint of Oxbow Books Published in the United Kingdom in 1996. Reprinted in 2015 by OXBOW BOOKS 10 Hythe Bridge Street, Oxford OX1 2EW and in the United States by OXBOW BOOKS 908 Darby Road, Havertown, PA 19083 © The author David Gallop Paperback Edition: ISBN 978-0-85668-675-7 A CIP record for this book is available from the British Library All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical including photocopying, recording or by any information storage and retrieval system, without permission from the publisher in writing. For a complete list of Aris & Phillips titles, please contact: UNITED STATES OF AMERICA UNITED KINGDOM Oxbow Books Oxbow Books Telephone (01865) 241249 Telephone (800) 791-9354 Fax (01865) 794449 Fax (610) 853-9146 Email: [email protected] Email: [email protected] www.oxbowbooks.com www.casemateacademic.com/oxbow Oxbow Books is part of the Casemate Group

ARISTOTLE ON SLEEP AND DREAMS

Reproduced, by permission of the University of Amsterdam, from ἃ sixteenthcentury miniature, which evidently illustrates the story about Aristotle (Diogenes Laertius V. 16) referred to in the Introduction, p. 34.

Contents ACKNOWLEDGMENTS PREFACE NOTE ON THE TEXT AND TRANSLATION ABBREVIATIONS INTRODUCTION 1. Aristotle's Life and Work 2

The Parva Naturalia

3 4.

The Greek Concept of Dreaming Plato's Legacy

5.

The De Somno and the De Insomniis

6 7

Dreams and the Imagination Dreaming and Teleology

8.

The Function of Dreams

9. 10. 11. 12.

Aristotle and Freud Dreams and the 'Daemonic Aristotle and Malcolm Conclusion

TEXT AND TRANSLATION On Sleep and Waking On Dreams On Divination through Sleep NOTES

APPENDIX

On Sleep and Waking On Dreams On Divination through Sleep Aristotle's Historia Animalium, IV. 10

GLOSSARY SELECT BIBLIOGRAPHY INDEX

Acknowledgments This book was first published in 1991 by Broadview Press in Peterborough, Ontario. I am grateful to Broadview for their cooperation in making this revised and corrected edition possible, and to the new publishers for welcoming it into the present series. The. book is not greatly altered in substance from the original edition, but it now includes, as an appendix, a

text and translation of Aristotle's Historia Animalium, Book IV, ch. 10, which

deals with the same subject-matter as his three essays on sleep and dreams in the Parva Naturalia. A new section 1 has been added to the introduction. The bibliography has been updated as indicated in the bibliographical note. A number of mistakes pointed out by reviewers of the first edition have been corrected, and several notes have been revised extensively in the light of suggestions from reviewers and readers. I am especially indebted to Professor Malcolm Willcock for the care with which he has commented upon both the original book and the revised version in draft. It is also a pleasure to thank Adrian Phillips for patiently generating a new text of a complex Work, and for his kind hospitality when I visited Warminster. I am grateful to Trent University for the sabbatical leave granted me in 1984—85, during which the original project first took shape, and for an extension of that leave into 1985-86,

which

was made possible.by a leave

fellowship from the Social Sciences and Humanities Research Council of Canada. From January to August 1985 I received support from a visiting fellowship in the Humanities Research Centre, and a teaching appointment in the Department of Philosophy, at the Australian National University, Canberra. I also spent the New Zealand spring terms of 1984 and 1987 as a visitor at the Universities of Auckland and Canterbury. Throughout these visitorships I received the most generous help and hospitality from my several hosts. Auckland University gave me office space and the use of its library when the project was getting under way. Canterbury provided the same support, and catered for all my word-processing needs, when the book was reaching its final stages. The Humanities Research Centre and the Philosophy Department at the A.N.U. afforded ideal working conditions, a wonderfully tranquil and congenial environment, and expert stenographic services. I also benefited from briefer visits to many other universities in Canada, the U.S.A., Australia and New Zealand, at which papers based on drafts of material in this volume were presented. My cordial

ACKNOWLEDGMENTS

vii

thanks go to all those departments of philosophy and classics whose members gave my thinking stimulus and direction. The introductory essay contains some material, especially in sections 7-10, that overlaps with a paper published in the Proceedings of the Boston Area Colloquium in Ancient Philosophy, volume IV (1988), which was itself based on a lecture given at Boston College in April 1988. Portions of that material were also used in the Gilbert Ryle lectures delivered at Trent University in November 1986. I am grateful to the BACAP organizers and to my own department for allowing this material to appear in its present form. More friends than 1 can mention individually helped me to wrestle with problems presented by this or that passage in Aristotle's text. I should like to express special gratitude, however, to two old friends and former colleagues at the University of Toronto: John Hunter, who read the whole translation and introduction in draft, and Joseph Owens, who read my entire manuscript. Their characteristically careful and probing comments helped me to improve the work a great deal. The original manuscript was prepared with capable assistance from Real Fillion. It also benefited in final revision from helpful suggestions made by a reader for the Canadian Federation for the Humanities. Faults and errors remaining are, of course, my own. David Gallop

Trent University, Peterborough, Ontario.

vill

Preface This book is designed for contemporary study of Aristotle's three essays on sleep and dreams. Considering the perennial interest of their subject-matter, these works have received surprisingly little attention in modern times. Even if the physiology of sleep and dreams is now understood better than Aristotle could have dreamt himself, his philosophical arguments and psychological insights on these topics read as freshly as ever, yet they remain largely unknown. The three essays form a single sequence and seem to have been written as such. I had originally intended to prepare a briefly annotated translation of them, and to deal in a separate volume with the larger historical and philosophical issues they raise. As the work grew, however, I found it hard to keep the two projects distinct, and eventually decided to merge them by setting the essays in their historical context, and making them the focus of a rather wider study. As a result, the book is distinctly different from many other recent presentations of Aristotelian texts. In no way does it pretend to the status of a full-scale critical edition or scholarly commentary. But by providing a modern format for material which might have been thought too arcane to be worth serious attention, I hope it will show that such texts can have lasting

interest and value. I have aimed to present Aristotle's ideas in a form that will be useful to non-specialists as well as to experts, and have therefore striven to avoid the commonest and deadliest sin that besets philosophical scholars, namely, Jargon. The introduction and notes consider the material not only from a philosophical perspective but also with reference to neo-Freudian interpretations of Aristotle and contemporary experimental psychology of dreaming. They may therefore hold some interest for psychologists and historians of science. Aristotle's scientific theories, and the array of observations he adduces to support them, display his usual marvellous boldness and ingenuity. In general, however, empirical questions have been given less attention here than theoretical or conceptual ones. In particular, little has been added to the commentaries of W.D. Ross (ed. Parva Naturalia, Oxford 1955) and P. Siwek (Aristotelis Parva Naturalia, Rome 1963) upon

passages of the De detail.

Somno and De

Insomniis dealing with physiological

PREFACE

ix

The editorial material will seem lengthy in proportion to so short a text. Aristotle's writing is, of course, densely compressed, and would have needed the sort of expansion that lecturers often give their notes in the classroom. But also, to serve the interests of different kinds of reader, my

format has

often led me to repeat points made in different contexts, and to signal these with cross-references. I have tried to restate Aristotle's arguments accurately, but without probing at all deeply into some of the most controversial aspects of his philosophy of mind and perception. These include the proper interpretation to be placed upon his soul-body 'hylemorphism'; the role of ‘common sense' (koine aisthesis) and its relation to the special senses; the relation between

perception and imagination (phantasia); and the place of the present essays in Aristotle's development. These intricate issues, the traditional locus of philosophical interest in the Parva Naturalia, have been the subjects of vigorous debate in recent years. To have delved far into them would have required a detailed and technical discussion of many other texts besides those presented here, an enormous task in which my original design could have become totally submerged. I have therefore touched only lightly upon them, doing little more than summarize such broad consensus as may be said to have emerged from recent scholarship. I have often referred the reader to those studies which have most illuminated these issues for me, particularly those by Charles Kahn, Malcolm Lowe, Martha Nussbaum and Malcolm Schofield. I gratefully acknowledge my debt to those scholars, and to many others on whose work I have drawn. Mention should also be made of Deborah Modrak's Aristotle: The Power of Perception (Chicago, 1987), of which I should have taken fuller account had my own project not been virtually complete when it appeared. References to modern works are given by author's surname and, where necessary, by the date of the relevant item in the select bibliography. The bibliography contains full details of all works cited. No knowledge of Greek has been assumed in the introduction or the main notes. All Greek words have been transliterated except in the textual notes. In the interest of readability, I have avoided a faddish literalism in translation. My version stays as close to the Greek as English idiom allows, yet contains about three English words for every two Greek ones. This fact reflects not only Aristotle's elliptical wording and pithy style, but also the incommensurablity of Greek with English. Much of the misfit is conceptual. Aristotle's terms often lack English equivalents, and some of them cannot be translated uniformly in every context. Sometimes they are used technically,

X

PREFACE

as explained in his other writings. All problematic terms occurring in the present works are listed in the glossary, both in Greek and in English, with brief explanations and with references to fuller accounts given elsewhere in the corpus. By far the hardest problem of translation is posed by the word-group that includes phantasia and its cognates, phantasma and phantastikon. A transliteration has been added in parenthesis at each occurrence of these words, after whatever English term has seemed best suited to the context. But the reader should bear in mind that many of these passages contain difficult problems of interpretation. The use of 'imagination', 'appearance', 'fantasy', or their cognates, is not meant to prejudge these questions. The English words are best read as mere stand-ins for phantasia or related expressions. | Many subtler misfits are less easily signalled. Notoriously, it is hard to capture the nuances marked by Greek particles, vital as these often are for an ' exact rendering of the argument. Frequently, too, the exact force of Aristotle's prepositions is uncertain. On these and many other points any translation will prejudge interpretation to an extent that it cannot acknowledge. All commentators on Aristotelian texts must be uneasily conscious of how many problems they are sweeping under the rug. Sometimes they can plead that Aristotle swept them there himself. I cannot make that plea for all of the places. where an adequate understanding still eludes me. But I hope that this study will encourage others to give these fascinating texts the attention they

. deserve and repay.

Note on the Text

and Translation

The Parva Naturalia is extant in fifty manuscripts, of which the oldest is the Paris manuscript E, dating from the tenth century. Indirect evidence for the text of the three present essays exists in paraphrases by Michael of Ephesus (c. 1090) and Sophonias (c. 1300). The manuscripts and their relationship to the indirect evidence are discussed in the editions of De Somno (1943) and De Insomniis et De Divinatione per Somnum (1947) by Drossaart Lulofs, and of the Parva Naturalia by Mugnier (1953), Ross (1955), and Siwek (1963). In very few places do variant readings within the present essays affect the sense significantly. The few manuscripts mentioned in the present edition are listed under 'Sigla' below. By permission of the Oxford University Press, the present Greek text is based upon that of Ross, with minor deviations at 454a26, 455b25, and 46302, and correction of misprints at 454b14, 456b15, and 460220. The deviations, and certain other places where the text or translation remains doubtful, are indicated.in the textual notes printed below the Greek and keyed to the translation. Variant manuscript readings mentioned in these notes are derived from Ross's critical apparatus, which may be consulted by those interested. Greek words excised by Ross have been omitted. Greek words appearing between angle brackets, « >, which were supplied by Ross, have been retained and translated. Ross's parentheses have been removed at 454a23-24, 455a17-22, 455a24-26, 455b5-8, 455b15-16, 458a30-32, 463b14-15, and 463b25-26. Except where indicated, the Greek has been construed according to Ross's punctuation, but Aristotle's sentences have often been broken into shorter ones, with periods instead of commas

or semicolons.

To mark off distinct

points in Aristotle's exposition, Ross's paragraphs have often been divided into smaller units in translation, with periods sometimes replacing weaker stops in the Greek text. As a further guide, brief summaries attempt to chart the argument of the paragraphs they precede. These summaries do not adhere strictly to Aristotle's wording, and often contain a measure of interpretation that goes beyond the text. Each is preceded by a reference number linking the. passage it covers to the relevant section of the main notes. The marginal numbers and letters in the text and translation are those of Ross's text. They derive ultimately from the edition of Immanuel Bekker (Berlin, 1831), whose page numbers provide the standard modern notation for references to Aristotle's writings. Each Bekker page number is followed

xii

NOTE ON TEXT AND TRANSLATION

by the letter 'a' or 'b'. Numbers following the letter, normally running from 1 to about 35, refer to the relevant line in Ross's text. For the three present essays the Bekker pages run continuously from 453b11 to 464b18. References in the notes to places within this passage are usually given by page and line number only. Line references are based on the line numbers in Ross's text, which are given, every five lines, in the margin of the present Greek text. The exact points of transition to a new Ross line-number are marked by vertical strokes throughout the Greek and the translation. Cross-references to the textual notes are given in the form ‘see textual note 15'. Cross-references to the main notes are given in the form ‘see note 44 on 462b12—26'. References in the form 'b12-16' are to lines within the section covered by the note in which they appear. References to 'Ross' are to his edition of the Parva Naturalia unless otherwise indicated. References to 'Beare' and 'Hett' are to the versions of 1.1. Beare in the Oxford Translation (Volume IIT) and W.S. Hett in the Loeb Classical Library. A revised version of Beare's translation, based upon Ross's text, appears in Volume I of the Complete Works of Aristotle, ed. J. Barnes. Where this version differs from Beare's, reference is made to 'Barnes'.

references for these and other sources are given in the bibliography. SIGLA E L M P S U

= Parisinus 1853 = Vaticanus 253 = Urbinas 37 = Vaticanus 1339 = Laurentianus 81.1 = Vaticanus 260 —

saec. saec. saec. saec. saec. saec.

x xiv xiv xiv xii-xiii xi

Full

xii

Abbreviations used for works of Aristotle An. Post.

Top. Phys. GC DA PN Sens. DM DS DI DD HA PA GA Probl.

Posterior Analytics Topics Physics De Generatione et Corruptione De Anima Parva Naturalia De Sensu De Memoria De Somno De Insomniis

De Divinatione per Somnum Historia Animalium De Partibus Animalium De Generatione Animalium

Metaph.

Problemata Metaphysics

EN EE Rhet. Poet.

Nicomachean Ethics Eudemian Ethics Rhetoric Poetics

References are made to the Oxford Classical Text, except in the case of EE, GC, HA, PA, and Probl.

Loeb Classical Library.

For these works references are given to texts in the

"Ihe priests’, I continued, ‘took to Aristotle generations ago, and have held by his teachings in a most striking manner.

For Aristotle's mind is much like a corkscrew,

being tortuous but powerful, and opening up worthy things for our satisfaction. reputation has surprised me somewhat, seeing how often he is wrong’. Douglas Woodruff, Plato's American Republic (New York, 1926)

His

Introduction 1. Aristotle's Life and Work

Aristotle lived from 384 to 322 B.C. His dates, and a rough chronology of his life, can be derived from Diogenes Laertius' Lives of the Philosophers (third century A.D.), which is our fullest source of information about him. His birthplace, Stagira, was a small Greek township close to Macedon, the powerful northern neighbour whose armies would overrun the Greek citystates and destroy their autonomy during his lifetime. His father, Nicomachus, was a doctor, the friend and physician of Amyntas II, king of Macedon.

His mother, Phaestis, came from Chalcis, on the island of Euboea,

where her family held property. Aristotle is likely to have acquired his lifelong interest in medicine and biology from his father. His Macedonian connections were also to have a profound influence upon his career. When he was about seventeen, he left home to pursue his education in Athens. There he entered Plato's Academy, the research school founded by the great Athenian philosopher some two decades earlier. He remained a member of the school, first as a student and later as a junior colleague, until Plato's death in 347. He then left Athens, and made his life for the next five years on the eastern side of the Aegean, engaged in scientific studies. His friend Hermeias, a close associate from his days in the Academy, had become ruler of Assos in Mysia. Aristotle settled in Assos, and married Hermeias' niece and adopted daughter, Pythias. She bore him a daughter of the same name, and may also have borne his son, Nicomachus. We are told that he remained in Assos for three years. Hermeias was then overthrown, and later cruelly put to death. A poem which Aristotle composed in memory of his father-in-law has been preserved by Diogenes Laertius (V. 7-8). From Assos he moved to Mitylene on the island of Lesbos, where his studies included research in marine biology. In 342 he was invited by Philip, who had succeeded as king of Macedon, to oversee the education of his fourteen-year old son, Alexander, the future world conqueror. Little is known of their association, which does not appear to have made any significant mark upon Alexander's career. Any intimacy between them seems to have diminished after 340, when the young man became regent for his father.

2

INTRODUCTION

In 335, after Alexander's accession to the throne, Aristotle returned once more to Athens, but he never rejoined the Academy. Instead, he formed his own circle at the Lyceum, the site of a public gymnasium, northeast of the city. It soon flourished as an independent school, which became known as the 'Peripatos', from the colonnade or covered walk where its gatherings were first held. Hence its members later came to be called 'Peripatetics'. The legend, transmitted by Diogenes Laertius (V. 2), that Aristotle or his successors were so named because they lectured "while walking up and down' is unfounded. Aristotle remained at the Lyceum for the next twelve years. At some point during this period his wife died, and a woman of his household named Herpyllis thereafter became his companion, or possibly his second wife. According to one dubious tradition, it was she who bore his son, Nicomachus, from whom the Nicomachean ethical treatises, derived its name.

Ethics, one of Aristotle's extant

The sudden death of Alexander in 323 was followed by an eruption of anti-Macedonian hostility in Athens. Aristotle, whose Macedonian connections were well known, found himself facing prosecution on a charge of impiety. Rather than stand trial, he took refuge in Chalcis, where he died in the following year at the age of sixty-two. He had retired, so he is said to have remarked, ‘in order that the Athenians might not commit a crime against philosophy for the second time’ — that is, by reenacting the judicial murder of Plato's master, Socrates. The remark, whether authentic or not, seems in character with the wry wit occasionally displayed in Aristotle's writings, and also suggested by various bon-mots attributed to him in Diogenes Laertius (V. 17-21). The same source has also preserved the text of his will, which testifies to his provident and generous nature (V. 11—16). Aristotle's writings, massive in bulk and encyclopaedic in scope, represent a monumental achievement. Diogenes Laertius (V. 22-27) lists more than 150 items, many of which have been lost. The extant corpus comprises about thirty treatises. Although these alone fill twelve volumes of the modern Oxford Translation, it has been estimated (Barnes

1995, 9) that they amount

to less than a third of Aristotle's total output. They evidence his activity both as a scientist and as a philosopher, ranging widely over biology. physics, astronomy, psychology, metaphysics and theology, besides exploring the basic concepts and principles of scientific inquiry. They contain epochmaking discoveries in logic, including the first system of principles for formally valid inference. They also contain fundamental contributions to moral and political thought, as well as seminal works on rhetoric and poetry,

INTRODUCTION

3

which have profoundly influenced the theory and practice of the arts. For nearly two millennia the world-view expounded in the Aristotelian corpus was to shape and dominate European thought in every field. Although that view has everywhere long been superseded, Aristotle's works have continued to excite the admiration of philosophers for their powerful insight, subtlety, and rigour of argument. Collectively, they stand as the greatest contribution to dispassionate inquiry made by any ancient thinker. They are, in a certain sense, as timeless as philosophy itself. Their style and format, however, present a considerable challenge. The extant treatises seldom display the literary qualities for which Aristotie was noted in antiquity. They can hardly be said to exemplify the ‘golden stream' of eloquence, for which he was praised by the great Roman stylist, Cicero. Rather, they belong to the category of writings known as ‘esoteric’. This term refers to ‘internal’ or intramural works, in contrast with those that Aristotle himself calls 'the exoteric discourses’. The latter were popular writings, which presumably account for his literary reputation, but which are extant only in fragments. Thus, most of our treatises are likely to have been generated for use within the Lyceum. They are generally austere in manner, tersely phrased, often elliptical or allusive, occasionally rambling, inconclusive or repetitious, and not always grammatically coherent. The rougher sections read like sketchy memoranda used in lecturing, or perhaps notes taken by a scribe or student. In many passages a distinctly professorial voice can be heard.

There are a few references to visual diagrams, and other

pedagogical allusions suggesting classroom delivery. A persuasive attempt has been made to reconstruct the equipment of Aristotle's classroom from such allusions (see Jackson). It has recently been urged that the treatises should be viewed as compilations by a later editor of Aristotle's ‘working drafts’ (Barnes 1995, 14-15). This too is highly probable, but is surely not incompatible with the "lecture-note' hypothesis. Even in our own schools, the distinction between 'lecture-notes' and ‘working drafts’ is not a sharp one. In the Lyceum too, we may assume, 'teaching' and 'research' were inseparable aspects of a single, shared enterprise. Whatever the genesis of Aristotle's extant works, they place unusual demands upon the modern reader. They best reward those who are willing to engage with them slowly. patiently, and in detail. This fully applies to the texts in the present volume, to which we may now turn.

4

INTRODUCTION

2. The Parva Naturalia

Aristotle's three essays on sleep and dreams (De Somno et Vigilia, De Insomniis and De Divinatione per Somnum)! belong to the collection of short treatises known,

since the late thirteenth century,

as the Parva Naturalia.

Within this collection our essays have traditionally occupied the third, fourth, and fifth places, being preceded

by works

on perception

(De Sensu

et

Sensibilibus) and memory (De Memoria et Reminiscentia), and followed by those on length and shortness of life (De Longitudine et Brevitate Vitae), on

youth and old age (De Juventute et Senectute), on life and death (De Vita et Morte), and on respiration (De Respiratione).

The canonical arrangement of these writings may not represent the order of their composition. Nevertheless, the essays on sleep and dreams are closely linked in subject-matter with the two that precede them and with one another. At the opening of the De Somno (453b11—24) and in the closing lines of the De Divinatione (464b16—18) the three works are presented as a unified treatment of a group of inter-related problems. The essays on sleep and dreams, like the works on perception and memory, supplement Aristotle's account of perception in his work on 'the soul' (De Anima), a general introduction to his series of psychobiological lectures. The De Anima promises at 432b11—12 to discuss sleep and waking, the subject of the De Somno. Conversely, cross-references to the De Anima occur in the De Somno at 454212, 455a8, 455225. and at DI 459a15.

inferences

as to relative dates of composition

references,

which

could have been

added

can be drawn

No safe

from these

later by Aristotle or his editors.

The De Anima 15 itself a patchwork of lecture materials, some of which may have post-dated the Parva Naturalia. But the treatises of the Parva Naturalia as now arranged are evidently meant to be read as a continuous sequel to the De Anima. On one topic, of central importance for the De Insomniis, the De Anima is clearly presupposed by it. That topic is the concept of imagination (phantasia), discussed at length in De Anima HI. ch.3. Whatever their chronology, the treatises of the Parva Naturalia belong thematically between the De Anima and Aristotle's zoological lecture-

I

The traditional Latin equivalents for the Greek titles are given here in full. the first and third will be shortened to De Somno

(or DS) and

In references De Insomniis will sometimes be abbreviated to DI.

From now on

De Divinatione (or DD).

INTRODUCTION courses,

the De Partibus Animalium

and

5

the De Generatione Animalium?

They often enter into greater physiological detail than the De Anima, consistently with its view that most psychological phenomena are inseparable from their physical basis, and need to be studied in that context (DA 403a10— b16). It should be borne in mind that the scope of these essays is far wider than the human species. Although Aristotle often writes with the human case in mind, he seeks in the De Somno an account of sleep that will cover animals in general. Thus at 455b31-34 he generalizes from human physiology to that of other animals. His discussion of dreams also mentions certain species other than man (463b12).

It has recently been shown that certain passages in the Parva Naturalia provide a resolution of difficulties in the De Anima, particularly as regards the relation between the special senses and the general sense-faculty of which they are separate modes or aspects.? It has been further claimed that the Parva Naturalia actually correct the De Anima in significant respects, and must therefore be among the latest of Aristotle's surviving works. While this latter claim remains controversial,‘ it is safe to say that the works on sleep and dreams represent an advanced phase of his thinking on psychobiological questions. The De Somno and the De Insomniis invoke the subtle views on the relation between soul and body and on perception that he was forging in the De Animas. The De Somno contains traces of a significant modification of his well-known doctrine of ‘four causes'5 The De Divinatione is implicitly critical of his early views." Several passages in the essays show signs of incomplete revision. There is reason, therefore, to assign them to a relatively late stage in Aristotle's period as director of the Lyceum (335-323 B.C.), and 2

3

For a reconstruction of the sequence, see Kahn, especially 4-6 with n.12, 20-21 with n.59. Kahn has argued persuasively that, whatever their order of composition, 'the De Anima and the Parva Naturalia form a continuous exposition' (16). See Kahn, esp. 6-17.

4

The arguments of Block for this thesis have been criticized by Hardie and Kahn (17—22),

5 6 7

but Block's work effectively challenged the older view of Ross (1955) 11—17, (1961) 9— 10, that much of the PN represents a transitional stage in Aristotle's development between the dualistic psychology of his youth and the more sophisticated teaching of the De Anima. In this respect Kahn's study (17) has confirmed Block's conclusion, with the important result that the PN can now be credited with a high degree of philosophical sophistication in their own right. See Introduction, section 5, and notes 2 on 453b24-454a11, 6 on 455a12-b2, 21 on 458b3-9. See the cogent discussion by Lowe, and note 8 on 455b13~34. As will be argued in section 10 below.

6

INTRODUCTION

to suppose that, along with the De Anima and other biological writings, they still awaited final revision at the time of his death.’

3. The Greek Concept of Dreaming In the Parva Naturalia Aristotle's approach to sleep and dreams is scientific throughout. He shows himself fully emancipated from superstitions that had prevailed before him, and persisted for long after his time. Nevertheless, it is worth asking how far his treatment of the subject may have been influenced by prescientific beliefs, or by the ordinary language in which those beliefs were expressed. It will be particularly important to be aware of this, if the Greek concept of dreaming differed significantly from our own. In one way, fortunately, it did not. Greek words for 'dream' present no serious problem of translation. Unlike many central terms in Greek philosophy, they do not lack modern English equivalents. Greek philosophical usage was, nonetheless, conditioned by a view of dream experience that we do not entirely share, and that can be seen to have affected even Aristotle's enlightened outlook. The

OED

defines

a dream

as 'a train of thoughts,

images

or fancies

passing through the mind during sleep‘. In line with this, we tend to think of a dream as a series of images, or of mental occurrences, possessing sequential order and (often) narrative structure.

A dream, thus conceived, has temporal

duration, and may have several different phases. It takes a certain course, in which first one thing happens and then another. We should be more inclined to liken it to a motion picture than to a painting. Moreover, we should distinguish the dream as a whole from the individual people or objects that feature in it. We would not naturally use the word 'dream' to refer to a single item within the dream narrative. For example, we might dream of a royal visit, in which the Queen arrived by aeroplane. We should say that she arrived by aeroplane ‘in the dream'. But we should not speak of the Queen herself as being the dream. She would be merely a central figure in the episode or story that the noun 'dream' signifies. By contrast, the Greek nouns for 'dream' (onar, oneiros, enupnion) did not originally signify a dream episode. The dominant early use of oneiros seems to have been not 'dream narrative' but 'dream figure', a person whom we should think of as appearing in the dream, but whom we should not identify 8

Forthe chronology of Aristotle's biological writings, see also Nussbaum, 10-12.

INTRODUCTION

7

with the dream itself? This usage embodies a notion of the dream, especially frequent in Homer, as a god or an agent sent by a god, bearing a message to the dreamer. Thus, in the /liad (II. 1-34) Zeus sends 'wicked Dream' (oulos Oneiros) to Agamemnon, urging him to attack Troy. Here Oneiros functions

as the proper name of a malevolent emissary, who appears to the sleeper in disguise.

It is characteristic

of a dream,

converse

with the dreamer,

giving advice or commands.

example

occurs in Plato's Phaedo

thus understood,

(60e-61a),

that it can hold

A much

later

where Socrates tells of being

visited by a recurrent dream (enupnion) that orders him to practise the arts. This dream, like Agamemnon's, can speak, and can change its guise from one visit to another. In such places the dream is both objectified and personified. It is viewed not as a mere figment of the dreamer's imagination, but as having independent existence, and as no less perceptible than an object of waking vision. A dream is, indeed, ordinarily spoken of as ‘seen’. It is not merely a subjective experience, such as a sleeper is said, in English idiom, to 'have'. This concept of a dream appears to have long outlived the beliefs that first sustained it. According to Dodds, 'the language used by Greeks at all periods in describing dreams of all sorts appears to be suggested by a type of dream in which the dreamer is the passive recipient of an objective vision'.? The Greek verbs for 'dream' can be used with a direct object or a prepositional phrase specifying what the dream is ‘of or 'about'.! They can also be used intransitively, and so without reference to dream content. But unlike the English 'dream', they do not govern a 'that' clause. By contrast, our propositional construction enables a complex story to be told within its scope, the verb 'dream' being understood to govern everything in the narrative that follows it. This may incline us to think of dreaming as a mental activity or process, a subjective counterpart to the story told in a dream report. Greek 9

10 11

See Dodds, Homer only dreams with Dodds, 105. these places Kessels

104, 122, nn. 9-10. He notes that oneiros as 'dream experience' occurs in in the phrase en oneiro(i), used in four places. For the tendency to identify persons, cf. Kessels, 7-10, 174—85, 199. Aristotle speaks of 'seeing' dreams at 458b20, 462b1—4, 463b8, 464a4. In his verb has been translated ‘experience’.

(198—203)

has

noted

the

late emergence

of separate

verbs

(as distinct

from

periphrases) for 'dreaming' in Greek literature. They are first found in the Hippocratic treatises and Aristophanes. Kessels plausibly connects this fact with the long persistence of the old 'objective' conception of the dream. The emergence of these verbs, he suggests, marks a stage in which 'the experience of dreaming has become the activity of the dreamer himself; the dreamer is no longer a kind of (passive) spectator to what was sent to him from the outside' (203).



INTRODUCTION

theorists would have been less tempted in this way. The model suggested by their idiom was, rather, that of a mental apparition, a ghost in consciousness, a phantom deriving from the old dream-figure that their nouns for ‘dream’ had first denoted. These figures were commonly said to be experienced 'dreamwise' (onar), an adverbial construction in which dream awareness was opposed to waking (hupar). Dreams were objects, especially persons, 'seen' in a distinctive mode of vision.” This is not to say that Greek dream experience lacked the narrative structure and content that are often possessed by our own. Greek literary dreams yield nothing to ours in dramatic interest or incident. Nor was the subject of real dreams merely the passive observer of an apparition. Aristotle himself mentions the common type of dream in which we rehearse our own previous or intended actions (463a23-27). But he pays no attention to these when explaining dreams scientifically. He assumes, rather, that what has to be accounted for is an internal ‘appearance’ (phantasma), a phantom appearing before the sleeper. Long before Aristotle's day the belief that dreams were objective visions of the supernatural had been challenged by rationalist thinkers of the early fifth century. Xenophanes had attacked the credentials of divination in general.? Heraclitus (DK 22 fr. B89) had declared that in dreams each of us retreats into a private world. He thereby dismissed them as subjective experiences, as no part of that 'common' world that he enjoined his readers to follow (fr. B2).!^ An early naturalistic explanation of dreams occurs in Herodotus (VII. 16). Xerxes, king of Persia, has been contemplating the invasion of Greece, and has been urged in a recurring dream to undertake it. His uncle, Artabanus, advises

him to disregard the dream's

advice.

For, he counsels,

"whatever

thoughts occupy a man's mind in the daytime tend to pervade the visions of his dreams'. But when Artabanus puts this theory to the test by donning the

12

The basic meaning of the onar/hupar contrast is uncertain. It is not clear whether the original contrast was between dreaming and waking vision, or between two modes of dream

13 14

vision,

deceitful

and

veridical.

Whatever its original sense, the contrast in Plato, whereas Aristotle never uses it. See Lesher, esp. 6-17. For this interpretation, see Dodds, 118, in the proverbial phrase 'you are telling ‘you are telling me what I know only too anyone else, presumably, because I have

See

Kessels,

186-9,

van

Lieshout,

gained a distinctly literary flavour. See Glossary, s.v. enupnion.

41-4.

It is common

131, n.91. At a later date subjectivity is implied me my own dream' (Republic 563d2), meaning well already’. I know my own dream better than privileged access to it.

INTRODUCTION

9

king's clothes and sleeping on the royal bed, he is visited by the same dream himself, and is forced to accept its divine origin.

Even Artabanus still talks

of dreams 'wandering about among men" and superstition has the last laugh on him. But the idea that dreams often merely rehearse our waking preoccupations came to play an important part in scientific theorizing. It occurs in Hippocratic literature (On Regimen IV. 88), and is later powerfully elaborated, for both human and animal dreams, by Lucretius (IV. 962-1036).

The same idea is invoked by Cicero in his scathing critique of divination (De Divinatione II. 119-50, esp. 139-41). Aristotle belongs in this rationalist tradition. He gives no credence to divine emissaries or ghosts of the departed. He rejects religious explanation in favour of the view that dreams are 'daemonic' (463b14), a notion whose significance we shall discuss later. Nevertheless, a trace of the old 'dreamfigure’ concept may be detected in his approach. For what he feels called upon to explain is not the whole episode that we should call a dream, but simply the phantom item appearing to the sleeping person. It is this apparition that he takes to be the dream proper: 'an appearance (phantasma) in sleep is what we call a dream’ (459a19-20). Consequently, once he has differentiated the relevant kind of phantasma from various others (462a15— 31), having underpinned the distinction with a physiological explanation, he can regard his work as done. The effect of treating the dream proper as an inner apparition is evident almost from the start of the De Insomniis. In proving that dreaming is not the work of judgment (458b10—25), Aristotle distinguishes the dream itself from various thoughts we may be having at the time. Just as what we perceive is distinguishable from our thoughts about it, so a dream is distinguishable from what we are concurrently thinking 'over and above the appearances (para ta phantasmatay (458b18). This, he adds, will be apparent to anyone who makes an effort to remember on getting up. Yet if we try to follow this advice, we shall find a distinction between our dreams

and our concurrent

'thoughts' far from easy to draw. The dictionary definition of 'dream' allows, as we saw, for a train of thoughts. We should often be inclined to say that our thoughts were part of our dream, or that ‘in our dream' we thought thus and so. Our word 'dream' certainly does not mean 'apparition in sleep' or ‘mental image in sleep'. It is questionable, in fact, whether a dream, for us, need contain any imagery at all. Even if imagery is a striking feature of full15 16

Using the old ‘objective dream’ language, as Dodds (118) points out. In section 10 below.

10

INTRODUCTION

blown 'hallucinoid' dreams, its presence distinguishes them from experiences of a more 'thought-like' character, for which the word 'dream' does not seem wholly out of place." Nevertheless, the idea of a dream as a private apparition, an inner likeness of some real external person or object, dies hard. It is deeply ingrained in Greek thought, a residue from the old 'objective dream' conception, an This ghost still lingers in internalized remnant of the Homeric ghost. Aristotle's talk of 'appearances' (phantasmata), and it may be said to have haunted speculation on the subject ever since. To dream of Coriscus is to be presented with a phantasma of him. By giving the genetic history of that appearance, Aristotle takes himself to have explained what a dream is. His account is thus narrowly circumscribed. He shows no concern with the relation between various images that may be combined in what we should think of as a single dream. He does not ask how a narrative structure comes to be imposed upon those images. He does not discuss the role of verbal language in the generation of dream fictions. Nor does he ever consider the possibility of distortion in recall and narration.5 He simply specifies what kind of phantasma a dream is, by giving its distinctive organic cause. In a general way, this conception of his task may be attributed to the 'apparitional view of dreams entrenched in ordinary Greek thought. But it can also be understood, more specifically, as a critique of certain aspects of his heritage from Plato. Τὸ that heritage we may now turn.

17

Dreams with imagery are highly correlated with the phases of ‘rapid eye-movement' (or "REM’) sleep which occur several times each night in normal human sleep, and which are known to have physiological counterparts in many other species (see p. 35 below). The few dreams reported after non-REM sleep have been found to be of a more 'thought-like’ character, whereas most dreams occurring in REM-sleep are said to have a perceptually vivid or ‘hallucinoid’ quality. See, for example, Crick & Mitchison, 112, col.2, 113,

col.2. Since there are obvious difficulties of classification, these claims should be treated with caution. They suggest some reluctance on the part of the subjects who report them to call the 'thought-like' phenomena of non-REM sleep 'dreams'. But they also show that English usage allows no such clear-cut distinction between 'dreams' and 'thoughts' as Aristotle takes for granted.

18

The

belief of Ross

(267-8,

note on 458b3-25),

that 458b20—25

refers to the

'rationalizing' of dreams, is probably a misinterpretation. Ross's note is confused and not entirely consistent, but if by 'rationalizing' he means the retroactive falsification of dreams, I believe he is wide of the mark. For my own effort to understand this obscure passage, see note 23 on 458b15-25.

INTRODUCTION

11

4. Plato's Legacy References to dreams are frequent in Plato's dialogues.? Often they echo proverbial notions from popular belief or literary tradition. For example, the Charmides (173a) alludes to the famous Homeric contrast between the Gates of Ivory and the Gates of Horn? i.e. between deceitful and veridical dreams. In the Republic (443b-c), when justice has been revealed as the principle that 'each should perform its own function', Socrates can say that a dream has been fulfilled. The principle of 'one man, one job', which had figured earlier in the discussion, has proved 'an image (eidolon) of justice' (443c4—5). It was a prophetic dream, a god-sent prevision of truth that has now come to light. We need not take Plato to be committed to the ideas that underlie these passing allusions. But other passages strike a more serious note. A remarkable section of Republic IX prefigures the Freudian notion of a dream as 'wish-fulfilment'. In this passage (571b-572b, cf. 574d-e) murderous and incestuous dreams are said to express desires harboured by outwardly decent people without surfacing in their waking thoughts. In their dreams they do such deeds as are performed in waking life only by that paragon of wickedness, Plato's 'tyrannical man'.?! In other places Plato uses dream fantasy to characterize his own enterprise in the construction of an 'ideal state'. To call it a 'dream city', a mere theoretical or verbal construct, is to suggest that dream fantasies gratify wishes in compensation for their non-fulfilment in waking life. It is in these terms, for example, that Plato floats a proposal for the radical re-ordering of sexual roles, which he admits would be virtually impossible to implement in practice. Its utopian character is brought out by representing it as Socrates’ day-dream (Republic 458a, cf. 450d).? These passages do not, of course, embody a general theory to the effect that every dream is the expression of a wish. Still less do they anticipate the Freudian view that most dreams are disguised wish-fulfilments. But they trade upon the familiar idea that in dreams we often compensate for the defects of the world we live in. Dreams are, on this view, a mode of fiction.

19 20 21

22

Fora survey see Gallop (1971). Odyssey, XIX. 541 ff. Wish-fulfilment is implied also at Theaetetus 173d, where the unworldly detachment of philosophers shows in their taking no interest in political and social gatherings, 'even in their dreams”. For several related passages, see Gallop (1971), 196-8.

12

INTRODUCTION

There is hardly a trace of this idea in Aristotle. Nowhere does he suggest that dreams embody states of affairs that the subject would like to see realized, but cannot achieve in waking life.? He takes himself to have explained them sufficiently by tracing them to waking perception. Apart from noticing that our waking actions or intentions commonly feature in dreams (463a23-27), Aristotle barely recognizes any connection between our dreams and our emotional preoccupations. He nowhere treats dreams as fictions that we actively construct, rather than as mere presentations that we passively undergo. It is not surprising, then, that he has no theory of dream symbolism, no account of why dreams should feature the bizarre episodes or peculiar narrative structure that they often do. Indeed, he hardly thinks of dreams as narratives at all. His guiding models are not stories but illusions or hallucinations. He explains the relation of dreams to the false judgments they produce in us, but he does not explore their relation to fiction. In the passage cited above from Republic IX, Plato had described varieties of dream in terms of parts of the soul operating independently of one another. Thus in evil dreams, the soul's rational part is asleep, and its lower elements assume control. In the sleep of a good man, however, these elements are quiescent, having been gratified in moderation and laid to rest. His rational soul is thus enabled 'to look for and reach toward the perception of what he does not know,

be it past, present, or future’ (572a2-3), untroubled by the

lawless desires and visions characteristic of Here we have a sophisticated version of travels. In popular belief the dreamer's bodily prison, can roam through space or commune with the dead, relive its own past 23

24

the lower parts. the idea that in dreaming the soul soul, temporarily freed from its time. It may visit distant places, experience, or gain knowledge of

Far from suggesting that the restraints upon waking behaviour of decent people are abandoned in their dreams, Aristotle once remarks that the dreams of good people are better than those of ordinary ones (EN 1102b10-11). This may be merely a consequence of his view that dreams are often a replay of waking actions (DD 463a23-27). But whatever may lie behind it, there is no suggestion that dream actions stand in any sort of contrast with waking ones, let alone that dreams afford a release for impulses that are inhibited in waking life. One possible exception is the obscure passage 458b20—25, as interpreted by Nussbaum, 250, n.45. As she says, Aristotle there uses phantasma not for a representative dreamimage presented to a passive subject, but for one that ‘results from the animal's active thinking and interpreting. But it is not clear that this ‘thinking and interpreting’ takes the form of composing a fiction or even (as she suggests) painting a picture. In the only example that Aristotle gives, they involve, rather, the use of a mnemonic technique. See

note 23 on 458b15-25.

INTRODUCTION

13

the future? Plato's rational soul attains, during sleep, a comparable liberation to pursue the truth. In an early work, On Philosophy, Aristotle had embraced this idea himself. But it is totally absent from the De Insomniis and the De Divinatione, and his theory of sleep implicitly rules it out. For the De Somno opens with an argument to show that sleep and waking are functions proper to neither soul nor body alone, but are common to both (453b24—454a11). By calling them ‘common’, Aristotle means that the waking and sleeping states consist in both physical and psychological aspects of the animal to which they belong. For an animal to be awake or asleep is neither for its body alone nor for its soul alone to be in a certain state, but for it to be functional or non-functional in the various modes of perception or thought of which it is capable. To predicate 'asleep' or 'awake' of it is to say something about both its physical and its psychological condition. On this view, it would make no sense to suppose that the soul could remain awake while the body was asleep, or that one part of the soul could remain awake while other parts were asleep. For waking and sleeping do not inhere in the soul alone or in its parts. They are psychophysical states belonging to the organism as a whole. Aristotle's treatment of dreams 15 in accord with this view. To understand them we must study them as biological phenomena. Their nature, like that of sleep, is to be explained in psychophysical terms. Aristotle is thus delivered from the notion of dreaming as a kind of spiritual travel. The soul is not a separable substance capable of making forays outside the body. Nor is it a recipient of divine messages or commands. Aristotle entirely rejects the idea that dreams come from God (462b20-22, 463b13-14, 464a20—22).

Plato had criticized the Homeric version of this idea:

God does

not send deceitful dreams to men (Republic 382e-383a). But he had still attributed dreams to divine agency himself, as may be seen from the Timaeus and the Sophist. In the Timaeus

(phantasmata) movements,

25

26

(45e-46a)

of real after the

dreams

external fire within

are treated

objects, the eye

images has

as internal

produced

ceased

appearances

by

to cause

residual vision

of

Cf. Dodds (102): For normal men [dreaming] is the sole experience in which they escape the offensive and incomprehensible bondage of space and time’. Dualism is clearly formulated in the Hippocratic On Regimen (IV. 86). For the related idea that the experient subject in dreams is a permanent, 'deeper' self, see Dodds, 135, 156-7, nn.1, 3. fr.12a (OCT. p.79) = Sextus Empiricus, adv. Phys. 1.21.

14

INTRODUCTION

external objects. Here we should notice (a) that a physiological explanation of dreams is closely linked by Plato with an account of mirror images (46a— c); and (b) his insistence that an explanation of that type is compatible with dreams having a divine origin. For the rational soul's powers of insight are exercised through reflections perceived by the irrational soul on the smooth surface of the liver (71a-e). That is why they need symbolic interpretation by a waking intelligence, if they are to be correctly understood. Dreams, then, are provided by God specifically for the purpose of divination. This permits a teleological explanation for them in accord with the wider program of the Timaeus. An explanation of that kind is precisely what Aristotle will rule out (463b13-14).7 In the Sophist (266—267) there occurs a fourfold classification of objects, according as they are made by man or God, and according as they are real things or likenesses.

In this classification, dreams

are treated, along with

shadows and reflections, as likenesses made by God, whereas pictures are likenesses made by man. In a memorable phrase, a picture is said to be 'a man-made dream for waking eyes' (266c9). By implication, a dream is thought of as a God-made picture for sleeping eyes. In this passage, then, dreams

are assimilated to other natural simulacra, and put down

to divine

agency. In the De Divinatione Aristotle emphatically denies a divine provenance for dreams. Rather, 'they are daemonic. For nature is daemonic, but not divine' (463b14-15). The full implications of this remark will concern us later/? Here we need only note that Aristotle writes as if previous failure to observe an important distinction were now being criticized. Plato had used both terms in the Sophist (266b7, c5) with apparent indifference. But it looks as if Aristotle is here repairing a previous failure of his own. For in the Eudemian

Ethics (1248a30-b2),

he had advanced a view of divination that

the De Divinatione seems to be correcting. The idea that certain subjects have God-given powers of prophecy is now replaced by a naturalistic explanation for successful guess-work (463b15-22, 464a22—b5). Like much else in nature, correspondences between veridical dreams and the real events

they presage are uncanny, but they are not literally the work of God. On these points, then, Aristotle came to reshape his Platonic heritage. But one crucially important element from Plato survives in his account. This is

27

For teleological explanations of dreams, see sections 7 and 8 below.

28

See section 10 below.

INTRODUCTION

15

the treatment of a dream as an inner likeness or reflection. A key text for this idea is Republic 476c2-d4, which deserves to be quoted in full: 'As for the man who believes in beautiful things but not in the existence of Beauty itself, nor is able to follow one who leads him to the knowledge of it, do you

think

that he lives in a dream

(onar)

or in a waking

state

(hupar)? Consider: is this not dreaming, namely, whether asleep or awake, to think that a likeness is not a likeness but the reality which it resembles?' I certainly think that the man who does this is dreaming. "Ihe man

who,

on the contrary,

believes that there is such a thing as

Beauty itself, who can see both it and the things which share in it, and does not confuse the two, does he seem to you to live in a waking state or in a dream?’ 'In a waking state certainly’, he said.

[tr. after Grube] Here the subject has before him a likeness of some real original. ‘Dreaming’ consists in his taking it to be not merely a likeness of, but actually identical with, that original. By defining dreaming in that way, Plato can use it to represent the condition of non-philosophers, who take sensible items for the transcendent realities, or 'Forms', of which they are likenesses. That is the point of the important addition, "whether asleep or awake’ (476c5—6).? Aristotle rejected the doctrine of transcendent Forms, along with the Platonic contrast between 'knowledge' and ‘belief, for which the 'dreamingwaking' contrast had been a symbol. He could have made no literal sense of the idea that someone might be dreaming while awake. Given the ordinary understanding of ‘dream’ from which he starts (459a19-20), the very notion

of a "waking dream’ would be self-contradictory. Although a definition of hope as 'the dream of a waking man' is attributed to him by Diogenes Laertius (V. 18), it was, if genuine, surely uttered in jest. But he himself uses a formula strongly reminiscent of Plato's, when he says that in dreaming ‘what is like something is judged to be that very thing’ (461b29). This accounts for the subject's supposing that a dream object is a real one. Like Plato, then, Aristotle explains the dreamer's error as a case of misidentification. The dreaming subject mistakes a mere likeness, i.e. a dream image, for a genuine sense-impression, and thus believes himself to be perceiving a real thing. 29

For the fuil implications of this point, and the wider significance of the dreaming-waking contrast in the Republic, see Gallop (1971), 190-4.

16

INTRODUCTION

The kind of ‘likeness’ used most naturally in this connection is a reflection. It is to this model that Aristotle repeatedly appeals. Thus, he can compare the bizarre contents of our wilder dreams with disturbed reflections in troubled water. Grotesque dreams are, as it were, distorted images of what we have experienced in waking perception. The distortions are attributed to internal ‘movements’ (kineseis), analogous to those of flowing liquids (461a8-25). At the end of De Divinatione (464b5-16) Aristotle compares a skilled interpreter of dreams with someone clever enough to reconstruct the original object, man or horse or whatever, that has cast a reflection upon troubled water. This comparison is of particular interest, for it highlights the defects of the ‘reflection model’ (as we may call it) with special clarity. Someone looking at a reflection of a horse in troubled water might mistake it for a reflection of a man. But no one could be said to mistake a dream of a horse for a dream of a man. For we have no criterion for establishing what a dream is ‘of independently of the dreamer's own say-so. Yet only by reference to some independent criterion could a subject who declared it to be a dream of a man be pronounced mistaken. The failure of the reflection model shows that (for us at least) a dream of X is not merely an internal image of X produced by the subject's waking perception of it. A dream of X does not stand to X as a reflection of X stands to X. The content of a dream need not be determined by its causal ancestry. In this respect Plato's analogy with a picture (Sophist 266c9) provides a better model. For it reminds us that the dreamer's own say-so plays a crucial role in determining what is dreamt, just as a painter's say-so plays a special role in determining what 15 depicted. Dreamers may be said, in some sense, to ordain the content of their dreams. We shall return to this point later.” Plato had raised one further issue regarding dreams, about which Aristotle says very little. In the Theaetetus (157e-158e) Socrates is made to examine and to combat the subjectivist doctrine of Protagoras, that all appearances are true for the subject to whom they occur. He mentions dreams, along with illusions, hallucinations, and delusions of the insane, as a prima facie objection to this doctrine. He then meets the objection by raising the question whether he and Theaetetus are awake or asleep, and by taking the exact correspondence between dreaming and waking appearances to throw doubt upon the matter. The implication is that dreaming, so far from being an objection to the Protagorean doctrine, may be seen as a further illustration of it. The lack of an acceptable criterion whereby to declare waking 30

In section 11 below.

INTRODUCTION

17

appearances more ‘true’ than dreaming ones suggests that both may be regarded as equally ‘true’ for their respective subjects. The appearances presented to dreaming subjects are true for them. From this passage it is clear that arguments from dreaming were already well-worn tricks of the subjectivist philosopher's trade in Plato's day. They are not relevant to Aristotle's interests in the Parva Naturalia. But in the Metaphysics (1010b3—11) he attacks an argument from dreams in the course of a polemic against the Protagoreans: ‘Next, one may fairly be surprised that they should find perplexing the question whether magnitudes and colours are such as they appear to those who are at a distance or to those who are near, and whether they are such as they appear to the healthy or to the sick; or whether what appears to the weak or to the strong is heavier; or whether what appears to the sleeping or to the waking is true. For it is obvious that they do not really consider it so: at any rate if someone in Libya believes himself one night in Athens, he does not set off for the Odeon’.

Here Aristotle argues that those who profess perplexity as to whether dreaming or waking appearances are true must be insincere. For they are not in any real doubt about the matter. The point of his argument is somewhat obscure. If he means that the Protagoreans do not believe that dream appearances are true, the behaviour of someone asleep and dreaming in Libya has no obvious bearing on the matter. But if what is at issue is whether dreamers themselves take dream appearances to be true, the Libyan example would not prove the contrary. Of course, no one sets off for the Odeon in Athens, while asleep in Libya. Being asleep in Libya (or anywhere else) would preclude any such action — except in sleep-walking, which is not relevant here. Probably, Aristotle means that one would not set off for the Odeon upon waking up in Libya and realizing that one had been dreaming. But that would show only lack of belief in the truth of one's dream appearances once they were already past, i.e. of dream appearances subsequently recognized as such. It would not show that current dream appearances are not true for the dreamer (as Socrates had suggested) while they are taking place. As for Socrates' question how we can prove that we are awake rather than asleep, Aristotle gives the matter short shrift, dismissing the issue as incapable of proof (Metaph. 1011a3-13). To demand a proof that one is awake 1s to make the mistake of supposing that everything admits of proof. Thus, he never considers dreams from the perspective that has engaged - -

18

INTRODUCTION

much philosophical attention since Descartes. He does not seek criteria whereby to distinguish waking from sleeping appearances, or to prove conclusively that one is not asleep and dreaming here and now. This indicates a significant point of contrast between ancient and modern philosophy. It has been well observed that ‘Greek philosophy does not know the problem of proving in a general way the existence of an external world'?! Ancient philosophers were preoccupied, rather, with finding a criterion whereby the true nature of that world might be discovered. Accordingly, dreams were invoked by them not, in the manner of Descartes, to question our knowledge of the external world's existence, but rather to question whether and how we can obtain any objective knowledge of whatever world there may be. For if all observation is inescapably conditioned by the observer, then the Heraclitean dictum that in dreams we enter a private world would apply equally to waking experience. In that case the Heraclitean injunction 'to follow what we have in common' could not be obeyed. All worlds would be private, and we would share no ‘common’ world at all. That is the subjectivist bogey so powerfully conjured up by Plato in the Theaetetus. Aristotle shows no sign of being troubled by the bogey, or of believing that dreams lend it any support. Rather, he takes dreaming to be a sort of ‘misperception’, and thus to be inherently deceptive in a way that ordinary perception is not. Thus at Metaphysics 1024b23-26 he uses dreams, along with shadowpictures (skiagraphia), as examples of 'things called false either because they themselves are not, or because the impression (phantasia) that results from them is of what is not’. [tr. Nussbaum, 247]. Nussbaum has correctly noted that Aristotle is here speaking ‘of all dreams, not just dreams about non-existent things or situations'# His stock examples of dreams are, indeed, of real items, e.g.

a man or a horse (458b10—

15) or Coriscus (461b22-462a8). But in accounting for dreams, he pays little heed to any distinction between those of existent and those of non-existent things or situations. He assumes, rather, that all dreams are of ‘non-existent 31 32 33

Burnyeat, 19. The implications of this point for the dream argument are well brought out in this article, 35-7. Αἱ 458032 he follows Plato (Theaetetus 157e3-4) in using 'mis-see' (paroran) and 'mishear' (parakouein). Plato had also used the generic 'misperceive' (paraisthanesthai). loc.cit. As she points out, phantasia in this context should not be taken to mean a ‘dream image’, but rather the false impression that results from a dream, viz. the taking of a dream item or situation for a real one.

INTRODUCTION

19

things or situations’, in the sense that whenever we dream, something is merely conjured up (phantazetai), which has no objective counterpart 'out there', currently present to the dreamer. That is consistent with his account of dreams as due to a residue from sense-experience that is no longer occurring (460b1-3, b28—32, 461b21—29), and his insistence that faint awareness of a

real lamp or a barking dog does not count as a dream (462a27—28).* A dream may, of course, prove 'veridical' in the sense that it matches an event occurring at some time or place in the real world; and it is just such dreams that Aristotle considers in the De Divinatione. But no dream could be veridical in the sense that any such real event is being witnessed by the sleeper during his dream. That is why dreams, like shadow-pictures, can exemplify things that are inherently 'false', i.e. (in the context of Metaph. 1024b23-26) delusive. For it is part of Aristotle's concept of a dream, as it is of ours, that when one is dreaming of something, one cannot be concurrently

perceiving that thing, or indeed anything else, in the ordinary way (cf. 458b8-9, 458b33-459al). If we now turn to the De Somno and the De Insomniis, we shall see why this is so.

5. The De Somno

and the De Insomnüiis

The De Somno reflects Aristotle's mature philosophy of mind, the view of the soul-body relationship sometimes called 'hylemorphism'. According to this view, soul and body are not two independent substances, each capable of separate existence. Rather, they are inseparable aspects, the 'form' and the 'matter' respectively, of a single living thing. Thus, sleep belongs, as we have seen, neither to the soul nor to the body alone, but is a psychophysical condition

(453b24-454a11),

(454a11—455a3).

both

Psychologically,

necessary

and

universal

it is a temporary

in

animals

suspension

of their

perceptual powers, which enables those powers to be exercised once more when the animal wakes (454a26-454b9). Physically, it consists in the rising of hot matter after the intake of food, and its reverse flow downwards (456a30-456b28). The action of hot matter upon ingested food affects the heart, which 15 a master-organ whose functioning is necessary for each of the special senses to be able to work (455a27—b13). Consequently, when this 'primary sense-organ' is disabled, the animal's consciousness, i.e. the

34

On this point see also pp. 38-40 below.

20

INTRODUCTION

‘common sense’ of which its special senses are so many distinct modalities (455a12-26), is cut off. Aristotle thus defines sleep as a state of perceptual incapacitation, to be differentiated from other states of unconsciousness (e.g. fainting) by reference to its physical cause (455b2-13, 456b6—28). We understand what sleep is, when we understand what organic functions are affected, and from

. what specific bodily cause their suspension results. A corresponding account is given for waking, in terms of the physical causes of waking up (458a1025). The De Insomniis takes a similar approach to dreams. Aristotle, as we have seen, assumes a view of them implicit in ordinary language. They are 'appearances' (phantasmata) that occur during sleep. This view is underpinned with an account of the physical processes that give rise to them. They are explained as due to traces from waking perception, which linger unnoticed in the sense-organs for some time after the external stimuli have departed, and later become reactivated during sleep. The traces are characterized in terms like those used later by Hobbes, when he spoke of ‘decaying sense‘. Aristotle describes at length the process whereby ‘movements’ (kineseis) travel from the sense-organs towards the heart, and (under certain conditions) mislead the sleeper into supposing that he is perceiving real external objects (461b7-462a8). Throughout this account a dream is assumed to be a psychological phenomenon, that has to be understood by reference to its physical basis. This enables the particular kind of 'appearance' (phantasma) that we call a dream to be distinguished from other kinds of ‘appearance’ (462a15-31), just as a physiological account of sleep enables the loss of awareness that constitutes sleep to be marked off from other forms of unconsciousness (456b9—12, b17-24). Both treatises, then, assume Aristotle's hylemorphism. Both assume that the nature of the phenomenon in question is discovered by asking to what part of the animal it may be attributed. Both consider the physical causes from which the phenomenon stems; and both distinguish the phenomenon from other similar ones in terms of their own particular causes. Nevertheless, a certain tension may be observed to exist between them. For the account of sleep given in the De Somno creates some difficulty for understanding dreams in the terms that Aristotle takes over from ordinary belief. If dreaming is taken to be a mode of perceptual awareness during 35

| Leviathan, ch. 2.

INTRODUCTION

21

sleep, yet sleep is defined as a state of perceptual incapacitation, the very notion of a dream seems threatened by self-contradiction. The De Somno says that a dream is 'in a certain way a sense-impression' (456a26). Yet it also explicates sleep as a state in which all ordinary perceptual awareness is suspended. If sleep is a state of perceptual unconsciousness, then how can we be conscious of anything during sleep, dreams included? This difficulty is in evidence from the outset of the De Insomniis. For the first chapter does not introduce the subject in the way that the program mapped out at the start of the De Somno (453b17—20) would lead us to expect, i.e. by asking what a dream is and what is its cause. Such an opening occurs only at the start of the second chapter (459a23-24), which seems better suited to follow directly upon the De Somno. Instead, the first chapter begins (458a33-b2) by asking what part of the soul is affected in dreams, and argues that they must be the work neither of perception nor of judgment, but of the imagination. We need not infer that the first chapter constituted a separate and later treatise.? But the De Insomniis, like the De Somno, shows signs of incomplete revision.* It seems possible, therefore, that the first chapter was drafted to deal with a difficulty that came to exercise Aristotle only after he had written chapters two and three, and that it had not been fully integrated with them at the time of his death. In that case, the chapter will represent an attempt to enable the causal account of dreams given in chapters two and three to be tied in to a solution of this difficulty. In the next section we shall follow the argument of De Insomniis ch.1 in further detail, in order to see how far the problem was resolved.

6. Dreams and the Imagination The De Insomniis begins by asking, in effect, just what mode of awareness dreaming is. Is it the work of the perceptual or the intellectual part of the 36 37

38

As Drossaart Lulofs (1947), xxix pointed out. As did Drossaart Lulofs. The same editor held that the De Somno likewise consisted of two separate treatises. This has not found favour with more recent editors, and has been decisively refuted by Lowe, 279-83. But the incoherences that Lulofs detected in both the De Somno and the De Insomniisare real. For example, 462a3]-b11 is tacked on to the end of the essay. Although it bears upon the questions raised at De Somno 453b18-20, no place has been found for it in the main exposition. Again, 459b23-460a32 coheres poorly with what follows it. For incoherences in the De Somno, see Lowe, 282-3.

22

INTRODUCTION

soul? Should it be viewed as a mode of perception or of thought? Aristotle argues that it is neither (458b3—25). Rather, a dream has to be understood as the work of the imagination (to phantastikon). To dream is not to be perceiving anything in the ordinary way, for in sleep the senses are not functioning: we are not literally seeing, hearing etc. anything at all (458b3-9). But neither can a dream be the work of judgment. For a dream, like an object of waking perception, needs to be distinguished from any judgment we make about it, e.g. our identification of what is presented to us as a man or horse, as pale or beautiful. Such judgments require perception, which is not, while we are asleep, occurring (458b10-15). Moreover, when we try to remember a dream, we can distinguish the dream appearance (phantasma) from thoughts we had while it was occurring. The dream proper was the appearance before us, and this is not to be identified with what we were thinking about it (458b15-20). Nor, as will become clear later (461b7-462a8), is it to be identified with the mistaken judgment, while we are asleep, that we are perceiving something real. Such mistakes are, Aristotle thinks, frequently made while we are dreaming, and he will later explain how they occur. But they are distinct from the appearances that give rise to them. Dreams are not, then, straightforwardly the work of either perception or judgment. Yet a clue to their origin, Aristotle suggests, may be found in perceptual delusions of the sick or the illusions of the healthy. In such cases of 'misperception' we are aware of something, but the object is not what we suppose it to be. Perhaps, then, dreaming is seeing or hearing something, though not in the ordinary way. The sense organs are affected somehow, though not as they are by external objects when we are awake (458b3045925). Instead, as Aristotle will show in chs. 2-3, they are affected by traces from waking perception, internal movements that produce 'appearances' (phantasmata) within us. Dreams are thus a sort of replay of previous waking experience, sometimes bizarrely scrambled as a result of physiological disturbance (461a14-25). In this way, Aristotle suggests, it is possible to understand dreaming as a kind of awareness during sleep, while still holding that all normal sensory functioning is suspended. Unfortunately, this solution is expressed in terms that are hard to interpret. Aristotle's concept of 'imagination' (phantasia) is difficult in itself, and its relation to perception (aisthesis) is particularly elusive. Despite his initial argument distinguishing dreaming from perception, he wishes to hold that dreams are, in a way, the work of our perceptual faculty after all.

For, he

says, 'the imagining part (to phantastikon) is the same as the perceptual (to

INTRODUCTION

23

aisthetikon), yet they differ in their being’ (459a15-17). "The imagining part' and ‘the perceptual’ are, then, alternative designations of one and the same capacity, having a common physical basis. A single capacity is describable both as 'perceptual' and as 'imagining', but a different account of its operation is required, according as it is viewed in one of those ways or the other.? Dreams are the work of this capacity, but they belong to it in its ‘imagining’ role, not its 'perceptual' one: 'it is plain that dreaming is the work of the perceptual part but belongs to this part in its imagining capacity’ (459a21—

22). What does this mean? How is this capacity supposed to function in its ‘imagining’ role? Here we are referred (459a14-15) to Aristotle's extended discussion of imagination in De Anima III. ch.3. But that discussion is itself full of difficulty, and it 1s not clear that a unified account can be extracted from it. Sometimes the capacity seems to be one whereby we have mental (especially visual) imagery, our ability to visualize sensible objects in their absence. Other passages suggest a more wide-ranging capacity whereby things ‘appear’ to us or strike us in a certain way. This will include, prominently, the capacity for 'seeing as', that is in play whenever visual phenomena are interpreted in a particular manner. In

connection

with

dreams,

it may

seem

natural,

at

first

sight,

to

understand ‘imagination’ in the first of these ways. For when Aristotle talks of dream 'appearances' (phantasmata), it is plausible to suppose that he means dream images, and is assigning these to a capacity whereby they are produced. Moreover, he calls imagination (phantasia) 'the movement that is produced by a sense in the course of its active functioning’ (459a17-18). This repeats a formulation from the De Anima (42921), suggesting that imagination's distinctive role consists in the production of imagery by the mechanism of 'decaying sense’. But this cannot be the whole story about ‘imagination’ as it pertains to dreams. For, in the first place, such a capacity could not provide a common factor that would justify treating dreams alongside various mistakes and illusions in waking perception which have no obvious connection with mental imagery. For example, 'the sun appears only one foot across, and yet frequently something else contradicts the appearance' (460b18—20, cf. 458b28-29). Or again, a single stick, when felt between crossed fingers, 39

Fora full discussion of the claim that imagination (to phantastikon) and the sense-faculty (to aisthetikon) 'are the same yet differ in being', see Nussbaum, 234-7, 255-61. See also note 3 on 454al1-b9.

24

INTRODUCTION

appears to be two, but no mental image of two sticks need be formed at that time. It simply feels as if we were touching two sticks: 'a single object appears as two' (460b20-21). Or the land seems to be moving to those who are sailing past it (460b26—27). In such cases it would seem neither plausible nor relevant to introduce mental imagery. Again, Aristotle's appeal to the imagination is presumably meant to explain dream images. But it would be vacuous

merely

to attribute

them,

after the fashion

of Moliére's

learned

doctor, to an 'image-producing faculty’. Moreover, even if Aristotle does mean to explain the genesis of dream images in that way, he also means to explain how we come to think, while dreaming, that we are perceiving real things. For that purpose an appeal to an image-producing faculty, or to the physical mechanism by which it works, would be insufficient. It is better, then, to interpret 'imagination' not simply as a capacity for mental imagery, but as one whereby something 'appears to', or is interpreted by, an observer in a certain way. At its broadest, it will be a capacity that enters into all of a creature's perceptual and cognitive engagement with the world around it. It determines 'how an object or state of affairs registers with the creature, what content it has for him', 'how things in the world appear to living creatures, what the creatures see their objects as’. In the present context, it is the capacity whereby, for example, people who are emotionally excited may see a stranger as someone they wish or fear to see (460b3-11), or feverish patients may assimilate cracks in the walls of their room to animals (460b11—16). Or again, it is the capacity whereby we may liken a cloud-formation to a man or a centaur (461b19—21). In dreaming, likewise, we assimilate an internally generated 'appearance' of Coriscus to an ordinary sense-impression of him, and we often think, accordingly, that we are seeing Coriscus himself (461b21—30).

Exercises of this capacity, as Aristotle emphasises, are compatible with differing states of belief as to whether the items in question really are what they appear to be (461b5—7, 461b30-462a8). The man who sees a stranger as his enemy believes that the stranger is his enemy, whereas those who liken a cloud to a man or a centaur do not for a moment believe that it is one. Patients may or may not think that the cracks in their walls are animals, depending on the severity of their illness. If we perceive something indistinctly, we will say that it 'appears to be a man', to register uncertainty as to whether it really is one or not (DA 428a12). Here we make no firm judgment on the matter. Accordingly, Aristotle distinguishes ‘imagination’ 40

Nussbaum, 246, 255.

INTRODUCTION

25

(phantasia) from ‘judgment’ (doxa), which may either endorse or oppose imagination's deliverances, or which may do neither. In dreaming it simply fails to oppose them, so that the appearances presented to the subject gain acceptance by default (461b29—462a8, cf. 459a6-8, 461b3-7). Aristotle is suggesting, then, that the capacity at work in dreaming 1s a wide-ranging ability to interpret phenomena, to see things in this way or that, to assimilate them to things which we may or may not believe them to be.^! But it remains far from clear that this conception of the imagination enables him to solve the problem that he invokes it to deal with in De Insomniis ch.1. For if our perceptual apparatus is incapacitated during sleep, as he has shown in the De Somno, it remains hard to understand how we can be aware of anything at all. If the heart, as the 'master organ' controlling the senses, does not permit ordinary perception, how can it permit a quasi-perceptual awareness of the internal 'movements' to which dreams are traced? Such awareness would seem no easier to understand, merely because it is endogenous, than awareness due to stimuli from an external source. Aristotle purports to solve the problem by assigning dreams to the perceptual capacity ‘in its imagining role’. But if perception and imagination have a common organic foundation, as their identification at 459a15-16 suggests, and if the exercise of this single capacity in its 'imagining' role requires, as does its performance in a 'perceptual' role, the functioning of a certain bodily organ, viz. the heart, how does Aristotle's solution help? If the effect of sleep 15 to disable the senses, should it not disable the imagination as well? Aristotle's frequent parallels between dreams and illusions tend to distract our attention from this problem. For in illusions, since the subject is not asleep, no shut-down of the perceptual apparatus occurs, and no comparable difficulty arises.

4]

The domain of the imagination, as we think of it, extends far beyond the kinds of phenomena mainly discussed in the De Insomniis. It includes the use of simile and metaphor, poetic imagery and symbolism, fable and myth, argument by analogy, mathematical and scientific models, and examples in philosophy. In all these 'imaginative' activities, we are assimilating one thing to another, viewing one thing in terms of or through the medium of another. In such activities, the capacity engaged seems comparable with that whereby we dream, as Plato intimated when he ascribed dreaming vision to mathematicians (Republic 533b-c), and when he discussed the use of philosophical examples in terms of the contrast between dreaming and waking (Statesman 277d, 278e). Aristotle, however, shows no interest in dreaming as a model for understanding other aspects of mental life, so the wider role of the imagination should not be over-emphasized in exposition of the DI.

26

INTRODUCTION

There are two further problems in the relation between dreaming and imagination.” At De Anima 427b16—24 Aristotle contrasts imagining with judging in two significant respects. (1) We can imagine things at will, whereas we cannot judge things to be the case at will: imagination is 'up to us when we wish (for it is possible to produce something before our eyes, as those do who set out things in mnemonic systems and form images of them); but making a judgment is not up to us, for one must either be wrong or attain truth' (trans. Hamlyn, adapted). (2) Our judgments as regards something terrible or encouraging are attended by the appropriate emotions: ‘when we judge that something is terrible or alarming we are immediately affected accordingly, and similarly if it is something encouraging; but in the case of the imagination we are just as if we saw the terrible or encouraging things in a picture. When we look at a picture, we are not possessed by terror, or perhaps by any emotion at all. Likewise, we may imagine something terrible, while remaining emotionally unmoved. Neither of these points fits the case of dreaming at all well. (1) It is not ‘up to us, whenever we wish' to dream or not to dream, to have some particular dream or not to have it, as we feel inclined. Dreaming is consistently treated by Aristotle as an 'affection' (pathos), something we undergo rather than something we do. Moreover, (2) a vivid dream of terrible events, far from leaving us emotionally unmoved, will often produce in us the most acute state of terror. We might, for example, imagine vividly that we have been kidnapped by terrorists, yet feel no fear at all. Our capacity to imagine it at will is reflected in the fact that we can, and often do, tell people to imagine or picture such things. The verbs 'imagine' and ‘picture’ have a use in the imperative mood. But the verb 'dream' has no such use. We could not sensibly be told to dream or not to dream of being kidnapped. Moreover, when we do dream it, we are quite likely to feel the fear appropriate to such an event in real life, not the emotional detachment with which we view a picture. So dreaming seems to satisfy neither of the conditions that Aristotle has placed upon the imagination. How, then, can he consistently attribute it to that faculty? One possible answer is that in the De Anima passage Aristotle is thinking of imagination narrowly as the conjuring up of mental images, since he there mentions the use of mnemonic techniques that require their formation. The points just made about volition and emotional detachment apply, then, only to imagination in the sense of image-formation, and not to all exercises of the 42

ἴῃ what follows I am indebted to Schofield, 123-5.

INTRODUCTION

27

imagination in the wider sense advocated above. Dreaming is an exercise of imagination tn this wider sense, and therefore lies beyond the intended scope of the two conditions. But he can also be defended in a different and perhaps more satisfactory way. Dreams are an admissible exception to both the De Anima generalizations, precisely because they occur during sleep. For in sleep the will and judgment are paralyzed, just as they are in fevered patients who take the cracks in their walls to be animals. Because of acute illness these patients cannot help thinking that the marks are animals, and may even behave as if they are. During sleep, likewise, our will and judgment are fettered, so that appearances are not subject to the control and critical appraisal that are possible when we are awake. Dream appearances are not subject to the will, because during sleep we usually lack any control whatever over what is presented to us. And because we assimilate dream appearances to real situations uncritically, we feel emotions appropriate to the relevant belief states in waking life. Thus, if we dream of being kidnapped, we are liable to feel terror, because while dreaming, we take the reality of the dream situation

for granted.

We believe, after a fashion, that we are being kidnapped.

In this

way, dreams, like the delusions of the feverish patient, can be attributed to the imagination, despite their failure to meet the two conditions that Aristotle

has placed upon the imagination in general. This defence can be reinforced by noticing a common experience in sleep to which Aristotle himself calls attention. He points out that illusory appearances are not endorsed by judgment, once the subject is aware of their cause. In this connection he observes: ‘if someone perceives that he is asleep, i.e. that it is a sleeping state in which the perception is occurring, then there is an appearance, but something in him says that it appears to be Coriscus, and yet Coriscus is not there. (For often something in the soul of a person asleep says that what is appearing is a dream)’ (462a2-7). This strikes many readers as well observed. It is also noticeable that any emotion we feel in a dream is instantly dispelled upon the realization that we are 'only' dreaming, and that what has been presented to us is not real. This strongly suggests that the feeling of emotion during a dream requires temporary belief, or at least suspension of disbelief, in the products of our imagination. This suspension is possible because judgment is disabled while we are fast asleep. As soon as we become aware of being (or having been) asleep, disbelief in what we have been imagining is restored, and our emotions immediately become disengaged. We no longer, as it were, believe the 'story' that we have been telling ourselves. Thereafter our attitude to

28

INTRODUCTION

events in that story becomes very different, if (as is perhaps seldom the case) the 'story-telling' process continues at all. We can, then, read Aristotle's attribution of dreams to imagination as consistent with the two points made about the imagination in the De Anima. For we can see how the special circumstances of sleep render dreams a legitimate exception to both those points. On the other hand, the fundamental difficulty noticed earlier still remains. If sleep involves a complete shutdown of our perceptual apparatus, it has still to be explained how we could have quasi-perceptual awareness during sleep, and how it could seem to us that we are seeing or hearing anything at all. Something more will be said on this question later.

7. Dreaming and Teleology We noticed earlier“ that Plato's account of dreams in the Timaeus (45e—46a, 71a-72b) makes them amenable to teleological explanation. By contrast, Aristotle's theory contains no teleology. It assigns to dreams no purpose, function, or meaning, and its physiological aspect has been deservedly labelled ‘an almost mechanical picture'.5 In this respect it stands in marked contrast

not only

with

Plato's

for example,

wrote:

ideas,

but with

those

of numerous

modern

theorists as well. Freud,

'

. Since

we

have

a habit

of looking

for

teleological explanations, we shall be more ready to accept theories which are bound up with the attribution of a function to dreaming'.# He maintained, as is well known, that dreams are the expression of a wish, often transformed by the dreamer's Unconscious. The true content of this wish is disguised, in order that it should not disturb the sleeper sufficiently to interrupt sleep: "When once we have recognized that the content of a dream is the representation of a fulfilled wish and that its obscurity is due to alterations in the repressed material made by the censorship, we shall no longer have any difficulty in discovering the function of dreams. It is commonly said that

43 44 45 46

Insection 11 below. Insection 4 above. Edel 162. Freud, IV. 75.

INTRODUCTION

29

sleep is disturbed by dreams: strangely enough we are led to a contrary view and must regard dreams as the guardians of sleep’* More recently, it has been proposed that 'the function of dream-sleep (more properly, rapid eye-movement or REM sleep) is to remove certain undesirable modes of interaction in networks of cells in the cerebral cortex'.48 The suggested function 15 to break certain synaptic connections, which if continuing unbroken might lead to impaired brain functioning in the waking state. The proponents of this theory go so far as to speculate that 'attempting to remember one's dreams should perhaps not be encouraged, because such remembering may help to retain patterns of thought which are better forgotten. These are the very patterns the organism was attempting to damp down'^ Other theorists have suggested that there is a functional link between rapid eye-movement sleep and animal learning, the former enabling the latter to take place more effectively.9 It has also been proposed, on evolutionary and neurological grounds, that human REM sleep may be a memory-processing mechanism inherited from the lower species, which enables individuals to develop an ongoing strategy for survival.5! If we look for any sort of teleological hypothesis from Aristotle, however,

we search in vain. teleological omission in Aristotle ‘is way, namely

conveyed'.*

One critic is surprised that he 'does not mention any

reason for the fact that dreams exist — surely an this teleologically minded thinker'.? According to not interested in the function of the dream, except in that dreams are not the means by which supernatural

interesting this critic, a negative wisdom is

In this and the following section, we shall see that Aristotle's

supposed 'omission' of teleology is both consistent with his general position and defensible in itself. Aristotle does, indeed, repudiate the traditional belief that dreams serve a supernatural purpose. They are not communications from God, he argues, or they would occur in the daytime and to more suitable recipients (462b20—22, 463b12-16). It need not surprise us that dreams are held to lack purpose in this popular sense. But given his usual approach to biological phenomena, we might expect him to hold that dreams are purposive or functional on a 47 48 49 50 51 52

Freud, V. 678. Crick and Mitchison, 111. op.cit. 114, col.3. There is a growing body of evidence to support this last hypothesis. See the article by Winson. Wijsenbeek-Wijler 216, cf. 234.

53

op. cit. 172.

See Smith.

30

INTRODUCTION

deeper level, more particularly if they are to conform to his own well-known fourfold pattern of scientific explanation. Should not dreams, like other organic phenomena, admit of the sort of explanation traditionally called a ‘final cause’? The critic just quoted slightly qualifies her remarks at one point. Aristotle, she allows, ‘does after all assign some function to dreams, namely, as symptoms, i.e. warning signs of an approaching physical disturbance'* Aristotle does indeed say that some dreams may be 'signs' of this sort (463a3-21, 30-31)55 In one sense of ‘meaning’, these dreams might be said to have one: they mean illness, much as clouds mean rain. But the use of dreams

in

medical

diagnosis

does

not,

in

Aristotle's

terms,

entail

the

assigning to them of a function or purpose. The fact that we may occasionally use X for a purpose P does not mean that P may be viewed as X's purpose or function. A sign of some future event enables us to predict that event, but this does not entail that what we used as a sign occurred for that purpose. The sun's entry into shadow is said to be a 'sign' of its imminent eclipse (462b30-31), but the ‘sign’ does not occur in order that we may predict the eclipse. Similarly, even if we may sometimes use dreams to diagnose or predict illnesses, that does not mean that they occur in order to warn us of illness or that such a warning role is their purpose or function. That is not what Aristotle's natural teleology is about. In general, a teleological explanation should, for Aristotle, conform to one or other

of two

models.”

According

to the

first, which

we

will

label

‘purposive’, some end or goal is specified that the agent consciously seeks to achieve. For example, a man may take a walk for the sake of his health. According to the second model, which we will label ‘functional’, something happens or comes into being or exists, 'for the sake of' some state of affairs, without its being the case that any subject is conscious of the end to which

54 55

56 57

op.cit. 237. Here Aristotle follows the tradition of Hippocratic medicine. Cf. Dodds, 119, 133, n.102. Since the Hippocratic author recognizes (On Regimen IV. 93) that dreams may be undisguised wish-fulfilments, it is all the more remarkable that Aristotle makes no reference to this idea. For the question whether human use of things in the natural world implies that those things are ordered for man's sake, see Wieland, 158. For these two kinds of teleological account, see Nussbaum, Interpretive Essay I, esp. 74— 88. For Aristotle's natural teleology more generally, see Ackrill, 41—54, Cooper, and Owens.

INTRODUCTION

31

the process leads or contributes. For example, leaves grow on plants for the sake of protecting their fruit (cf. Phys. 199a23—26, DA 412b1-3). It is clear that for Aristotle no explanation of dreams could be expected to conform to the first model. For two basic conditions requisite for purposive explanation do not obtain in their case. If an agent A does X for a purpose P, then (1) it must be the case that A can do X or not do X

as Aristotle would say, ‘up to A’.

at will, i.e. that X is,

Moreover, (2) A must expect that by the

doing of X, P will be achieved, often (in the case of a human

agent) on the

strength of a belief that X generally promotes P, e.g. that walking generally promotes health. It is evident that dreaming satisfies neither of these conditions. (1) As already noticed, we do not dream at will. Whether we dream at all, or what we dream, is not normally ‘up to us'. This accords with Aristotle's repeated characterization of dreaming as an ‘affection’ (pathos 458b2, b28, 459a9, 412, 464a19-20), with its implication that the dreamer is passive: dreaming is something that happens to him, or something he undergoes, rather than something he does. (2) We have no basis in experience for expecting dreams in general, or any particular kind of dream, to promote any specific outcome, hence for believing that by dreaming we shall achieve any particular purpose. Nor do most people hold such expectations or beliefs. For these reasons, the suggestion that someone had dreamt in order to achieve a conscious purpose makes no sense for us, and could hardly have made

more

sense to Aristotle.

It is unremarkable, then, that his account of

dreams should contain no explanation of a purposive kind. It would be surprising if it did. It might be objected that at least one type of dream does admit of purposive explanation, namely, the deliberately sought dream, which the recipient has tried to induce for a specific purpose, such as the cure of an illness. The use of various techniques, such as prayer, fasting, or incubation, in order to procure dreams, was common in antiquity, and is said to persist in some societies today.9* And what are such induced dreams if not purposive? It matters not whether they really achieve the ends desired. So long as the recipients believe or hope that they will, may they not be said to have dreamt in order to achieve those ends? Moreover, in cases where dreams contained

58

See Dodds, 110-16.

32

INTRODUCTION

sound advice, as they occasionally did a belief in their efficacy must have seemed justified. All that can be said here, however, is that such individuals have taken steps that are within their power, such as fasting, incubation etc., for the

purpose of procuring a dream, so that they may be cured. They have not dreamt in order to be cured, as one takes an aperient in order to induce a bowel movement. They have practised various techniques in order to obtain a dream, through which they hoped that a cure would be effected. We could say that they wished for a dream, but not that they dreamt at will or on purpose. They hoped that dreaming would serve their purpose, but the dream itself, as distinct from the steps taken to induce it, would not admit of purposive explanation. This way of viewing the matter accords with Aristotle's own account of purposive action, which links it with deliberation about how to achieve ends that we wish for but that are not directly within our power to realize. We deliberate, says Aristotle, not about our ends but about the things that promote them, and it is the latter that we may be said to choose, or adopt for a purpose: ‘our wish relates rather to the end, whereas choice is of things conducive to the end; thus, we wish to be healthy, but choose things whereby we shall be healthy' (EN 1111b26-28, cf. EE 122639-10). Accordingly, we might deliberate about whether or not to seek a dream, or about how to do so (e.g. which

of various techniques to use); whereas

we could not deliberate

about whether or not to have a dream, or whether to dream of X rather than of Y. We can no more choose to have a dream or not to do so than we can choose to be or not to be healthy, and for essentially similar reasons. In terms of this account of purposive action, then, even a deliberately sought dream would not be subject to the sort of purposive explanation that Aristotle could accept.

8. The Function of Dreams

If dreaming has no purpose, has it any function? Certainly, it is a functional rather than a purposive explanation that modern psychology has looked for, as can be seen from the quotations at the start of the previous section9 Moreover, it is just that sort of teleological account that we might expect to 59 60

op. cit. 115. See pp. 28-9 above.

INTRODUCTION

33

find in Aristotle. He has argued that dreams are the work of the imagination (to phantastikon). But if the operations of that faculty are the exercise of a capacity, like the exercise of the senses in perception, or of the intellect in thought, should its workings not be amenable to functional explanation? If dreams are amongst imagination’s products, may we not ask what function they fulfil? In this connection, a critic quoted earlier has drawn attention to the difference between Aristotle's treatment of dreams and his account of sleep. He is, she remarks, 'keenly interested in the function of sleep, and not only in its nature, as is to be expected in the framework of his teleological thinking. Yet he is not interested in the function of the dream ...“. How, it may be asked, is this discrepancy to be explained? It is instructive to compare Aristotle's theory of dreams with his account of sleep. For the asymmetry between them 1s not at all hard to understand. Indeed, from Aristotle's treatment of the latter we can see why no parallel account is available, or is even to be expected, for the former. In the De Somno Aristotle argues (455b13--28) that for animals possessed of perception and thought, sleep is necessary to enable them to exercise those capacities when they wake. Since no animal can exercise them continuously for more than a limited time, it must sleep periodically, so that it may resume its activities once more upon waking. Thus, indirectly, sleep serves to enable the animal to realize its nature as a mature member of its species, hence to exist as the kind of thing that it is. This thought is recapitulated at the end of the treatise: sleep occurs 'of necessity, since it is not possible for an animal to exist, should the conditions that produce it not obtain. Yet it is for the sake of animal preservation. For rest does preserve it' (458a29—32). Here Aristotle is applying to sleep the well-known doctrine of 'four causes' that he expounds and uses elsewhere. There is reason to believe that he had modified this methodology by the time he wrote the latest portions of the De Somno, and that he would no longer have talked in terms of a ‘final cause’ for sleep.” In the Metaphysics (1044b8-20) a lunar eclipse is given as an example of a phenomenon that need not be 'for the sake of' anything. The same would apply, by implication, to sleep, which is mentioned as a further example in the same passage. Nevertheless, the De Somno retains an argument to the effect that sleep provides animals with an intermission

61 62

Wijsenbeek-Wijler, 172. See Lowe, 286-7 and note 8 on 455b13—34.

34

INTRODUCTION

necessary for their continuing activity. That is why it must be universal in animals (454a26—b9). Elsewhere (DA 412a23-26, cf. 412b10-413a2) the relation between sleeping and waking provides Aristotle with a model for one of his central ideas, the distinction between the possession of a capacity and its active exercise.“ Waking is said to be analogous to studying, whereas sleep is analogous to the mere possession of the relevant capacity. When a geometer, for example, is not exercising his knowledge, the capacity for geometry lies, as we say, ‘dormant’ within him; whereas when he is studying or proving a theorem, it is actively engaged, 'awake'. Thus, when Aristotle identifies waking with ‘the end' (455b22-23), he thinks of a living creature using those perceptual and intellectual powers through whose exercise it achieves its well-being. This idea is nicely illustrated by an anecdote in Diogenes Laertius (V.16) about Aristotle himself. He had a bronze ball placed in his hand whenever he was lying in bed, with a pot set beneath it, so that if he dozed off, he might be woken by the sound of the ball dropping into the pot, and thus (presumably) enabled to exercise his own mind and senses as continuously as possible. If a functional explanation were available for dreams, it would have to show that dreaming promotes the self-realization, preservation, or well-being of the organism in a similar way. The creature's waking performance would have to be in some way enhanced. Dreaming could then be seen to enable it to exercise its capacity for perception and thought in waking life, and thus to be the kind of living thing that it is. That would provide a functional explanation of a truly Aristotelian kind. But could Aristotle possibly have believed, let alone shown, this to be the

case? Would it have been in the least plausible for him even to suspect that dreaming sharpens our perceptual or intellectual capacities, or that it is as necessary for their exercise as are regular periods of sleep? He had no evidence to suggest that human beings perceive or think better as a result of dreaming, let alone to suggest that without periods of dreaming they could hardly do so at all. Whereas it must have been as obvious to him as it is to anyone (even without the benefit of modern experiments in sleepdeprivation), that loss of sleep can be severely detrimental to perception and thought, and may even impair those capacities altogether.

63

On this and some other Aristotelian uses of the sleeping-waking contrast as a model, see Sprague (1977).

INTRODUCTION

35

With respect to sleep we find him not only arguing a priori that it must be universal in animals (454a26—b9, b23-27), but adducing empirical evidence

for it in every species except testaceans (454b14—22).

Further empirical

support for this claim is found in a chapter of the Historia Animalium IV. 10 (printed as an appendix to this volume). Dreams, by contrast, Aristotle does not take to be universal. He says that they occur in 'some' animals other than man (463b12), and elsewhere attributes them to all viviparous quadrupeds, specifically mentioning dogs, horses, cows, sheep, and goats (HA 536b2730). He is uncertain whether oviparous creatures dream, though it is obvious that they sleep (HA 536b30-32). Even in man, who is said to dream most of all the animals (HA 537b13), Aristotle holds that dreams do not occur at all in infants till they are four or five years old (HA 537b14-16, cf. 461a11—13,

Probl. 957a19-22); nor do they occur after the intake of food (461a11—14, cf. 462b4-8). He also notes, most significantly, rare cases of adult men and women who have never dreamt at all, or who begin to do so only late in life (462a31-b4, cf. HA 537b16-20). Yet he never once suggests that their perceptual or intellectual performance might be diminished as a result of their not dreaming or improved when they begin to do so. On the contrary, he reports that for such ‘late dreamers’ physical change has subsequently ensued leading to debility or death (HA 537b19-21). The foregoing observations might be thought only to show the primitive character of Aristotle's biology. For every one of the points just made can be contested in the light of contemporary sleep research. Rapid eye-movement sleep, which is highly correlated with human dream reports, has been found to occur not only in human beings and quadrupeds, but in many viviparous mammals, as well as in primitive marsupials such as the opossum Moreover, electro-encephalograph readings characteristic of REM sleep have been found in birds, and in many types of fish, reptile and amphibian The onset and duration of REM sleep does not appear to be related in any way to the intake of food. So far from not occurring in young infants, it actually takes up a far higher proportion of their sleeping time than it occupies in the sleep of older people.© It takes up still more time for the foetus in utero.” In general, it would seem, the younger the animal, the more REM sleep it

64 65 66 67

Crick Evans Crick Crick

and Mitchison, 112, col.2. 129-30. and Mitchison, 112, col.2. and Mitchison, loc. cit.

36

INTRODUCTION

needs to take.® As for Aristotle's non-dreamers, they would now appear to have been simply mistaken. For ever since the discovery of REM sleep, it could be said that 'all humans, even those who believed that they never dreamed, did dream and did so every night for a substantial part of the night. Even the most recalcitrant non-dreamer could be converted to this new point of view, for all one had to do was to persuade him to spend a night in the laboratory, wait for the first burst of REMs and wake him up'® Thus, we could assume that Aristotle's supposed non-dreamers had simply not remembered

their dreams.

Indeed,

Aristotle himself once

mentions

this

possibility (453b18—20), though it is not explicitly pursued further.” Moreover, there is no longer a total lack of evidence to suggest that dreaming can enhance performance in the waking state. It is well established that when animals and humans are deprived of REM sleep, and are then allowed to sleep normally, they subsequently 'compensate' for REM sleep loss by spending longer periods in the REM phase.? There is some evidence of adverse effects attending REM sleep deprivation, and it has been suspected, if not yet decisively proved, that prolonged deprivation will produce psychological disorder.” There is also experimental evidence that animal learning ability is correlated positively with the amount of REM sleep registered.” Must not Aristotle, if he were faced with such findings as these, recognize a function for dreaming?

For it now

seems likely that it does, after all, in

some way enhance the performance of the waking organism. If this proves to be the case, his theory will not only be discredited, but will fail even to provide the type of explanation required by his own program. At this point, however, we need to guard against semantic confusion. It is important to distinguish dreams, in the ordinary sense of the word, from the phenomena monitored in a psychologist's sleep laboratory. In ordinary usage, neither our word 'dream' nor Aristotle's equivalent word simply means the bodily processes that are highly correlated with subsequent reports of 68

Evans 156, 198.

69

Evans, 126.

70

71 72. 73

| Given his own explanation of dreams, Aristotle must have 'people always dream when asleep, but do not remember' as pp. 50-51 below. But it would have been perfectly consistent professing 'non-dreamers' have invariably and totally forgotten Crick and Mitchison, op.cit. 112, col.2, 113, col.3, Evans, 133. Crick and Mitchison, 113, col.3, Evans, 131-3. Listed and discussed by Smith. Cf. also Evans, 218-19.

regarded the theory that mistaken. See section 11, with this to conjecture that their dreams.

INTRODUCTION

37

dreaming. Even if there were a 100% correlation between dream reports and REM sleep phenomena, this would not make the terms ‘dream’ and 'REM sleep activity’ synonymous. Aristotle might, even then, still rightly refuse to identify a dream with its physical correlates. But in actual fact, the degree of correlation so far found to exist between dreaming and REM sleep phenomena is far from sufficient to establish their identity. Sometimes those phenomena occur, yet no dream is reported. Conversely, dreams are sometimes reported in the absence of the phenomena.” Scientific usage of 'dream' is at present in a state of confusion. Psychologists are often reluctant to redefine the verb operationally, so as to mean ‘exhibit REM sleep activity'. But even were they to do so, the fact that REM sleep phenomena have an organic function would by no means suffice to prove that ‘dreaming’, in the ordinary sense of the word, must have one as well. It seems not unlikely that the functions of REM sleep activity are performed just as effectively in subjects who remain unawakened and report no dreams at all as in those who wake up with awareness of having, in the

ordinary sense, ‘dreamt’. It has even been suggested that these functions are performed more effectively in the former subjects, on the hypothesis that awareness of having 'dreamt' occurs only when the organic processes of REM sleep have, for some reason, been aborted.^ Perhaps those processes, like those of digestion, perform their function best when we are least conscious of them. What we call 'dreams' would then be mere epiphenomena, like stomach rumbles during digestion. Aristotle, then, if he were confronted with the findings of recent research, might agree that REM sleep brain-activity probably does have an organic function (as yet poorly understood), yet he might still deny that dreams in his (and our) sense of the word need have a function. They could be mere 'sideeffects', lacking any function of their own at all. We may, indeed, credit him with just such a view.

For on his own theory,

dreams are produced by 'decaying sense', and are therefore mere by-products of waking perception. Perception itself, of course, has essential organic functions. It enables the animal to move about, adapt to its environment, find food, protect itself, and thus to survive and flourish (cf. DA 434a30-b27, De

74 75

Crick and Mitchison, 112, col.2. Crick and Mitchison, op.cit.112, col.2: ‘Since the majority of dreams are not remembered, [REM sleep's] function is more likely to be associated with the unconscious dreaming process — that is, with REM sleep without awakening — rather than with the few dreams which are recalled'.

38

INTRODUCTION

Sensu 436b10-437a17). But its traces that are activated during sleep may serve no needs at all. Aristotle is therefore no more obliged to assign a function to dreams than to after-images or other optical side-effects, which he thinks can be explained similarly in terms of 'decaying sense’ (459b5—23). From this perspective, moreover, we can see why a functional explanation for dreams is, for Aristotle, uncalled for.

For he treats them as comparable

with sensory illusions, and explains them by analogy with misperception (458b25-459a8). They are prone to mislead us, as long as they remain uncorrected by our judgment. Yet, in general, functional explanations cannot be provided for perceptual failures. We can provide a mechanistic explanation for failure of eyesight, but we could no more explain it in functional terms than we could explain the malfunction of (say) a camera in

those terms. Since cameras are made to go right, not to go wrong, their going wrong does not have any function. If dreaming is thought of as ‘misperception’, as some sort of perceptual aberration, then the type of explanation that Aristotle gives for it is the only one he could be expected to provide. A functional account should be neither sought nor expected.

9. Aristotle and Freud

We have already seen that Aristotle's account of dreams has little in common with Freud's. Unlike Freud, Aristotle treats the physical aetiology of dreams as central. He postulates no function for dreams. He pays no attention to dream symbolism, nor has he any concept of wish-fulfilment. Anticipations of Freud have nevertheless been seen in certain of his remarks. It will be of interest to examine a few of these in more detail. Near the end of the De Insomniis (462a19—25) Aristotle says that some people may be aware of external stimuli impinging on them while they are asleep, though faintly and as if from a distance. Thus, they may be dimly aware of a lamp light or of cocks crowing or dogs barking. On waking up, they may recognize the sources from which those impressions came. On the basis of these remarks, it has been claimed by a neo-Freudian interpreter that Aristotle well understood dreams of a certain type, namely, those 'elicited by the impact of sounds on the sleeper's sensorium; these the dream-work converts into the appropriate imagery and then incorporates into the dream, with relatively little distortion'. In this process, we are told, 'the 76

Devereux, 272, cf. 5, 30, 297, n.17.

INTRODUCTION

39

dreamer denies their external origin, by "properly" reacting dream; this dispenses him from waking up at once, in order these stimuli also in reality’. In this passage, then, Aristotle have pre-echoed the Freudian idea that dreams serve to protect This interpretation, however, is a distortion of the text.

cite these phenomena as examples of any type of dream. insists that they are not dreams

at all (462a18-19,

to them in the to respond to is supposed to sleep.”

Aristotle does not

On the contrary, he

a27-28).

They

do not

count as dreams for him, because the senses are functioning, at least feebly, and consequently the phenomena experienced are not remnants of previous sense-experience. Because they are produced by concurrent external stimuli, they do not fit Aristotle's own definition of dreams properly so called. Again, Aristotle does not say that the sleeper 'denies' the true origin of these experiences or that the external stimuli are converted by 'dream-work' into imagery'. He says that they impinge upon the subject in a faint or muted fashion.

These phenomena

are not, then, meant

to illustrate the situation

when external stimuli become woven into a dream, as when a ringing alarmclock is heard as a fire-engine, or a bed-rail falling on the sleeper's neck is felt as the knife of a guillotine. Aristotle's point is, rather, that not every appearance in sleep counts as a dream, because if the senses are operating, however feebly, then what is presented through them is not a dream at all. An appearance can be called a dream only to the extent that it does not stem from a concurrent external stimulus. It is, of course, possible that someone should dream of a barking dog, when there actually 1s a dog barking within earshot. Aristotle need not be taken to deny that. But he would say that if the barking sounds are actually impinging upon a sleeping person's sensorium, then that person is not dreaming of a barking dog but is actually hearing one. To the extent that we are affected by external stimuli, we are not dreaming but perceiving. It is, then, part of our concept of a dream that our dream awareness of X should be

distinguishable from any concurrent perception of X itself. This subtle conceptual point finds its place in Aristotle's formal definition of a dream a few lines later: ‘an appearance that arises from the movement of the senseimpressions, while one is in the sleeping state and in virtue of one's being

TI

See the quotations on pp. 28-9 above. Cf. also Freud, V. 681: ‘Every dream which occurs immediately before the sleeper is woken by a loud noise has made an attempt at explaining away the arousing stimulus by providing another explanation of it and has thus sought to prolong sleep, even if only for a moment’.

10

INTRODUCTION

asleep! (462a30-31). This passage, then, lends no support to a Freudian interpretation.’ There are, however, some remarks at the end of the De Divinatione, which have been more plausibly read along Freudian lines.” Aristotle there compares the skilled interpreter of dreams with someone who can reconstruct the original that has produced a distorted reflection in troubled water: 'the most skilled interpreter of dreams is one who can observe resemblances'

(464b5-7). This sounds more Freudian. For the art of dream-interpretation, both ancient and modern, has consisted in 'observing resemblances'. Are not analysts who tell their patients that a gun or a trumpet in their dreams 'represents' a phallus practising just such an art? Are they not decoding the cipher in which their patients' sexual preoccupations have been cast by their Unconscious, and thus discerning the true content of which the dream image is a distorted reflection? The Freudian comparison may seem all the more apt if we assimilate the product of the dreamer to that of the poet, and the task of the psychoanalyst to that of literary critic. In the Poetics (145925) Aristotle connects a gift for "metaphor' with the perception of resemblance'. It is, then, possible to see the same faculty at work in the having of dreams as in the poetic use of metaphor. Both are activities of the imagination in a broad sense, the capacity to assimilate items in our experience to other things which they resemble in some limited respect.®! Freud too regarded psychoanalytic dream interpretation as akin to literary criticism. Yet, when read in its context, our passage is not nearly as Freudian as it may have seemed. Aristotle 1s considering whether there is a genuine skill of divination based upon dreams. Aside from two kinds of dream that he regards as intelligibly precognitive, he is prepared to dismiss veridical

78

79 80 81

Αἱ 463all-17 Aristotle mentions several further kinds of stimuli impinging upon sleeping consciousness. These phenomena, unlike the sounds made by barking dogs, are misinterpreted by the sleeper, and might seem more amenable to Freudian treatment. But Aristotle does not call them 'dreams', nor could he do so consistently with his official 'decaying sense' theory. For the stimuli here are entirely endogenous, and are registered while they are occurring: they are not due to remnants of previous waking perception. In context, they serve not to illustrate dreams, but only to show how internal stimuli too small to be noticed by day may have a powerful impact at night. They were noted by Freud himself (IV, 97, n.2, cf. 319-20). Cf Dodds, 133, n.107. Cf.note 41 on p. 25 above.

INTRODUCTION

41

dreams as due to ‘coincidence’ (sumptoma, 463a31-b11). But then he has to explain various supposed precognitive powers of certain sorts of subject. Among these are unstable people (melancholikoi), who have an uncanny knack of making guesses that turn out to be correct. After suggesting (464a32-b5) that their 'powers' can be explained in terms of their unusually rapid and varied thought processes, he turns to a different kind of expert. This critic's judgment, by contrast with the irrational powers of 'melancholics', is informed by genuine expertise. That is why he is introduced as ‘most skilled' (technikotatos, 464b5-—6). In what does his skill consist? Presumably, he is like those medical practitioners whom Aristotle has commended (463a4—7) for saying that close attention should be paid to dreams, because they may be signs of incipient illness. The skilled dream critic, by tracing dreams back to their origin in waking perception, is enabled to diagnose the bodily condition that has produced them. It is in this light that we can understand the comparison with reconstruction of an object that has produced a distorted reflection in troubled water. A dream can manifest similar distortion. For, says Aristotle, 'direct dream-vision (euthuoneiria) is erased by the movement’ (464b16).

What is meant here by ‘direct dream-vision'? Translators have resorted to ‘clarity’ or 'vividness' of dream vision. But euthuoneiria should mean something more like ‘straight’ or 'direct' vision. When dreams are distorted, it is not clarity or vividness that is erased (for even the most grotesque dreams may be clear or vivid), but rather the original experience from which the dream is derived. That is what the dream-critic is recovering when tracing the ‘reflection’ to its source.

The 'movement'

(kinesis) which erases 'direct

dream-vision' is, presumably, the disturbance engendered by certain physical factors,

e.g. fever,

intoxication,

or heat due to the ingestion

of food

(cf.

461a13-25). If this is right, we can see that there is nothing very Freudian about this passage after all. For the 'resemblances' observed by the dream-critic are simply those between the dream appearances and whatever sense-remnants lingered in the sleeper's organs and caused those appearances. They are not, on

that view,

resemblances

between

dream

symbols

and

the

ideas

they

represent. They hold, rather, between dream appearances and things that cause them, i.e. ultimately, according to Aristotle's 'decaying sense’ theory, the things actually perceived by the dreamer in waking life. In that case, ‘interpretation’ of the dream need go no further than tracing its origin in the dreamer's conscious waking experience. ΑἹ] the critic need do is to discover the source of the dream in waking perception. He does not have to connect it

42 with

INTRODUCTION elements

in the subject's

emotional

life, or uncover

its unconscious

latent content. It is important to notice here that distortions are treated as if they were exceptional rather than normal, and as if their causes are physical, or (as in

the case of 'melancholics') traceable to temperamental factors, which themselves have a physical cause. Nowhere is it suggested that distortion of the dream serves the emotional or instinctual needs of the subject in any way. Aristotle's theory may at this point seem to encounter a difficulty that Freud's escapes. The ‘decaying sense’ theory will, it seems, account for only a limited

range

of dreams,

if we

assume

that sense-remnants

can persist

inside us for only a limited time# We can understand how, after spending the day with Coriscus, we could dream of him the next night. It 1s less easy to understand how we could dream of (say) our long dead great-aunt, whom we last saw fifty years ago, or of Queen Victoria, whom we never saw at all. In such cases, therefore, it is tempting to speculate that a residue from our waking perception of Coriscus gets misinterpreted as our great-aunt or as Queen Victoria, because of some prior emotional preoccupation. There is, however, nothing in Aristotle's text to suggest that he held any such theory. He does, in one important passage (460b3-11), draw an analogy between dreams and cases where people misinterpret what they perceive when awake because they are emotionally excited. But this is no more than analogy, designed to explain how the assimilation occurs. He never suggests that our emotional preoccupations are activated in sleep, or systematically 'disguise' our wishes in the way that Freud supposed. Any resemblance to Freudian dream interpretation is therefore extremely remote.#

82

83

84

Devereux is therefore quite unjustified in associating Aristotle with the idea that ‘the dream's springboard is the real waking experience of the previous day rearranged into an uncanny pattern, fitting the requirements of the irrational instincts’ (4, n.6, cf. 67, 73). Aristotle says nothing at all about ‘irrational instincts’. Aristotle does not say for how long decaying sense-remnants could persist. But could he suppose them to persist, even in a ‘latent’ state, indefinitely? It is not at all clear that his account of the way they become reactivated (461b11—30) could accommodate this possibility. Could residual 'movements' be stored in the blood for ever? Αἱ 463a21-31 Aristotle mentions a causal connection between our dreams and our past, present or intended future actions. The causality may run, as he notes, in either direction. His point is simply that in such cases it is intelligible that a dream should give foreknowledge of the subject's own actions. He thus makes a limited concession to the view that dreams afford a basis for predicting the future. There is no suggestion, as

INTRODUCTION

10.

43

Dreams and the 'Daemonic'

There remains one well-known, and extremely intriguing, Aristotelian remark about dreams to which Freud called attention. In the De Divinatione Aristotle says that although they are not sent by God, they are nevertheless ‘daemonic’. For, he adds, ‘nature is daemonic, but not divine’ (daimonia all ou theia, 463b14—15). On this Freud commented in 1900: "no doubt this

distinction has some great significance, if we knew how to translate it correctly'.5 His own later gloss upon it (1914) was as follows: Dreams 'do not arise from supernatural manifestations, but follow the laws of the human

spirit, though the latter, it is true, is akin to the divine’.% Freud was right to see significance in the distinction between 'daemonic' and ‘divine’. But did he understand the distinction correctly himself? He was certainly right in taking Aristotle to be rejecting the supernatural. Dreams are 'daemonic' in virtue of their belonging to the natural order, not as having a transcendent source. But Aristotle says nothing about laws of the human spirit' and knew no psychological 'laws' as Freud conceived of them. Nor does 'akin to the divine' suit the context as a way of understanding 'daemonic'. Although Aristotle regarded the human intellect as 'akin to the divine' (EN 1177a13-16, 1177b26-1178a8, 1179a22-29), he would have made no such claim for other aspects of our mentality explored by Freud. He has, indeed, just mentioned animal dreams in the same breath as human ones (463b12), to show that dreams in general are not divine.

What 'daemonic' does mean here can be seen from the immediate sequel. For Aristotle goes on to infer the 'daemonic' character of dreams from the fact that it is people of quite commonplace intellectual ability who have precognitive visions. Their humdrum mental capacity is a proof that 'veridical' dreams are 'daemonic' (463b15-22).8 Aristotle compares their success with that of lucky gamblers. Given a long enough run of throws, the

86

Devereux (296, 298, n.146) imagines, that dreams may cause or contain conscious or unconscious ‘acting out’. Dodds (120) endorses this comment, but without giving an interpretation of the Aristotelian remark himself. The tone of Freud's comment suggests that in 1900 he had no confidence that he understood the distinction between the divine and the 'daemonic'. In fact, he expressly disclaimed the expertise needed to study Aristotle in depth. Not until the edition of 1914 does he offer an interpretation of his own. Freud, IV. 2-3.

87

The claim that dreams are daemonic is no mere parenthesis, as Ross's text suggests, but is

85

the main point of the paragraph, supported by the argument of 463b15-22.

44

INTRODUCTION

chances are that some of them will be successful. That is what happens with supposedly 'precognitive' dreamers. Their visions may correspond to a remarkable degree with actual events. But that is no more than the long arm of coincidence would lead us to expect, given the unusually frequent and varied visions that occur to people of their mentality. In short, the 'daemonic', in this context, is the factor of chance. To call veridical dreams 'daemonic' is to treat them not as divine communications but as lucky coincidences. In what sense, then, is 'nature' said to be daemonic at 463b14? It might be supposed that the reference is not to nature in general, but simply anticipates the explanation of precognition in terms of the special 'nature' (463b17) that lucky dreamers possess: it is their 'nature', and not divine inspiration, that makes them so successful. Aristotle will then be merely rejecting the sort of

Platonic view that he had once held himself. We have suggested earlier® that this may be part of his meaning. But there is reason to suspect that 'nature' is said to be 'daemonic' with a more general point in view. For at the start of the De Somno, when formulating a program for the De Divinatione per Somnum, Aristotle had asked whether prevision of the future is possible, and if so whether it is limited to events due to human agency, or whether it includes 'things for which the daemonic has responsibility (aitia), i.e. events that happen by nature (phusei) or fortuitously (apo tautomatou) (453b24). Here things that happen 'by nature' are attributed, along with 'fortuitous' events, to 'the daemonic'.

Events

that occur

‘by nature'

are, then,

a subclass

of those

ascribed to the daemonic. Presumably, it is in virtue of their belonging to. that subclass that ‘nature’ is said to be 'daemonic' at 463b14. If this 1s correct, we may hope to shed light on the meaning of 'daemonic' in the later passage by asking what feature common to natural and fortuitous events would justify attributing both of them to the daemonic in the earlier one. An answer to this question suggests itself in the light of Aristotle's discussion of natural teleology and chance in Physics II (especially chs. 4—6). From that discussion we may say that natural and fortuitous events are alike in three respects: (1) they are not contrived by a designing intelligence; yet

88

89

The reference to 'people who play at odds and evens' (463b20) should be compared with Rhet. 1407b1-5, where this game is again used to disparage diviners. Their prophecies are less likely to be falsified, if given without date. Aristotle compares them with gamblers who guess merely 'odd' or 'even' rather than any particular number. See p. 14 above.

INTRODUCTION

45

also (2) they might have been so contrived; and (3) they look were. It is precisely this combination of features that makes lous. The daemonic, then, will be the appearance without intelligent design. In Physics U. ch.6 Aristotle gives an analysis of what it is occur fortuitously. Roughly, it is to say that the event, though had been brought about by design, was in fact accidental. One arrival of a riderless horse at a place of safety (197b15).

just as if they them marvelthe reality of for an event to looking as if it example is the

This was fortuitous,

in that it occurred without human agency, yet the horse might well have been guided to safety. Or again, a tripod just happened to land on its feet: ‘it stood for someone to sit on, but it did not fall for some one to sit on’ (197b18-19). It was not, but might easily have been, purposely so placed. Or a falling rock happened to hit someone, yet did not fall for that purpose. It fell fortuitously, ‘in that it might have fallen through someone's agency and for hitting’ (19763 1-32). A spectacular example of this last type occurs in the Poetics (1452a4—11). Aristotle there shows how 'the fortuitous' can provide suitable material for a good dramatic plot. He cites the case of a man who had encompassed the death of a certain Mitys, and who was himself killed by the statue of Mitys falling upon him, while he was looking at it® The statue did not fall in order to kill the murderer. Yet in a grimly ironical way, it gave the appearance of design. It was an event such as might well have been brought about by design, e.g. by the dead man himself, embodied in his own statue, and taking — like Don Giovanni's victim - a piquant style of revenge. Here we have an occurrence that is 'fortuitous' par excellence. No agent brought it about, yet it resembles a deed such as an avenging daemon, if there were such a thing, might have done. We might characterize it as 'uncanny', meaning that it is just as if there were an agent at work, when we know that there was not. With the word 'uncanny', we should rule out an agent explanation, by alluding to the resemblance between this event and deeds perpetrated by genuine agents?! The contrast intended in saying that nature, and therefore dreams, are 'daemonic but not divine' can now be better understood. The term 'divine' 90

ΟΥ̓ perhaps, though less appropriately for the interpretation suggested here, ‘while he was attending a festival’. See Lucas, 126-7.

9]

Aristotle

emphasizes

that

what

makes

fortuitous

incidents,

like

that

of the

statue,

‘marvellous’, and therefore dramatically effective, is precisely their appearance of design: 'the most marvellous even among fortuitous events are those which appear to have happened as it were on purpose’ (Poet. 1452a6~7).

46

INTRODUCTION

would imply that phenomena are deliberately produced by an intelligent agent, as Plato had supposed in the Timaeus and Sophist; that dreams are part of a nature that 1s itself cunningly fabricated by a Super-Craftsman. By contrast, to call dreams 'daemonic' is to deny that they are the workings of such an agent, while conceding the appearance of agency that they present. Likewise, by calling nature 'daemonic', Aristotle expresses marvel at its phenomena, yet rejects the inference that a designing agent is literally at work. It is just this kind of relation that Aristotle is suggesting between veridical dreams and the events in which they are ostensibly fulfilled. A supposed prevision of some event that actually transpires is uncanny. For it is just as if some agent had warned us of its imminence, or just as if we ourselves had direct awareness of the future. But nothing of the kind is really true. More generally, it is indeed uncanny that in sleep we should experience those strange simulacra of real life, those weirdly scrambled replays of waking vision, that provide such a wealth of material for painters, poets and story-tellers. They give us the impression, sometimes irresistibly, that we are being 'told' something. Yet if dreams are merely a fortuitous by-product of waking perception and its physical aftermath, can that impression itself be anything but a wondrous illusion? That is the line of thought suggested by calling dreams 'daemonic'. If it was Aristotle's, we can see how far from a Freudian outlook his perspective upon dreams really is. For to call them 'daemonic' is simply to deny that there is any agent at work producing them within us. They are merely curious side-effects of a physical process. The Aristotelian position, then, is precisely the one that Freud laboured so strenuously to combat: Traiime sind

Schaüme, ‘dreams are froth’. On this view Aristotle's failure to assign a ‘final cause’ to dreams is fully consistent with his natural teleology. Dreams have no function, no purpose at all.® Here we can see a radical contrast with Freud's concept of the Unconscious, in so far as that is treated as an agent with purposes of its own. Dreams are explained by Freud as serving the purposes of that agent, or of the dreamer in so far as he is under its control. But Aristotle recognizes no such thing as an unconscious purpose. To treat dreams as 'daemonic' is to deny a purpose to them, much as we should deny it to the statue of Mitys, which crushed his murderer, but which did not fall in order to do so. 92 93

Freud, IV. 133. Seesections 7 and 8 above.

INTRODUCTION

47

In a later passage Freud speaks of 'the "daemonic" power which produces the dream wish and which we find at work in our unconscious’. He does not refer to Aristotle in that context, but only to 'antiquity' in general?* But if he had Aristotle in mind, he had again misread him.

For, if we have understood

Aristotle aright, it was not his intention to reinterpret the old religious concept of a 'daemon' in terms of an unconscious ‘force’ at work within the human mind. Rather, the daemonic has, at this stage of his thought, cut loose from its religious roots. Gone is the daemon who protects and prospers a man in life, and escorts him to the next world at death. Gone are the daemones that haunt the world of classical tragedy, the forces that destroy Phaedra or Medea, that punish Hippolytus or drive Orestes mad. Gone is the mysterious inner voice (to daimonion) that had spoken to Socrates. Gone are the daemones mediating between man and god, whereby Diotima had shown Socrates the nature of love. Nowhere in view are the mischievous daemones that will eventually become the 'demons' of Christian mythology or the evil spirits of modern folk belief? Aristotle might here be said to have 'demythologized' the daemonic. The term is used to deny that natural phenomena, including dreams, are the expression of a divine will. Yet it also acknowledges that dreams will sometimes give every appearance of marvellous design. Aristotle's hypothesis that most 'veridical' dreams are coincidence was intended primarily to undermine the credentials of divination. If correspondences between dreams and future events are mostly coincidence, then they cannot be used by any genuine art of divination as a basis for prophecy. For any such art would require a grasp of regularities in nature, ‘constant conjunctions’ between dreams and the events they presage.% Yet 94

95

Freud, V. 614, 676: "The respect paid to dreams in antiquity is, however, based upon correct psychological insight and is the homage paid to the uncontrolled and indestructible forces in the human mind, to the "daemonic" power which produces the dream wish and which we find at work in our unconscious’; '... in the hypothesis which we have set up in order to explain the dreamwork a part is played by what might be described as a "daemonic" element. For these 'demons', see Dodds, 41-2. Also 127, n.53 and 134, n.112, where Dodds firmly rejects the idea that, in calling dreams 'daemonic', Aristotle was attributing them to

'daemones in the air’, On the view urged above, to be sure, nothing was further from his

mind. 96

Cicero

For an opposing view, see the article by Bos. (De

Divinatione

11.146)

fastens

upon

this

point.

He

dismisses

dream-

interpretation as so much pseudo-science, contrasting it with astronomy. The latter is based upon observed regularities, objective data of a sort that dreams, by their very nature, (so he thinks) cannot provide.

48

INTRODUCTION

‘coincidence’ by its very nature precludes this:

‘no coincidence happens

either always or for the most part' (463a2-3, b10-11).

Freud likewise did not

think that dreams could be used for predictive purposes. He regarded popular belief in their prophetic significance as mere wishful thinking.” But Aristotelian scepticism can be extended to other kinds of significance that are more congenial to post-Freudian thought. Some dreams strike us as having the properties of well designed pictures or stories. They invite the sort of interpretation that is appropriate to literary or artistic works. True, they are not subject to conscious control. But they seem like the productions of an ingeniously creative artist within us, a clever film director, who contrives them not according to ordinary artistic principles, but according to some bizarre canons of his own. Even among those who would reject the detail of Freud's theory, this way of regarding dreams remains a legacy of his teaching. The same goes for those not uncommon experiences in which we go to sleep with an unsolved problem weighing on our minds, and wake to find that a solution has been suggested to us by a dream. Elias Howe, inventor of the sewing-machine, is said to have dreamt of being captured by savages, who threatened to spear him to death unless he could produce a machine that would sew.

He failed; but as he awaited execution, he noticed that each of

their spears had, close to its tip, a hole resembling the eye of a needle. Did Howe's Unconscious solve his problem for him while he slept? Aristotle, if correctly interpreted above, would reject that inference. He would agree that dreams often display a startling quasi-artistic or inventive pattern. But that, he would say, is fortuitous. To ascribe them to a purposive agent called the Unconscious is to posit just the sort of explanation for them that his use of 'daemonic' was intended to rule out.

11.

Aristotle and Malcolm

Our discussion so far has taken for granted a traditional distinction between dream experiences and the bodily processes that provide their physical basis. For Aristotle, as we have seen, a dream is a certain sort of 'appearance' (phantasma) presented to certain subjects at certain times during sleep. To dream is to be the recipient of such a presentation. Given this understanding, 97

‘The popular mind is behaving here as it usually does: what it wishes, it believes' (V. 674).

INTRODUCTION

49

it is essential that subjects be aware of their dreams. They need not, of course, be aware that they are dreaming, for they will often fail to recognize their dreams as such. But they must be aware of something. The notion of an ‘unconscious dream’, a dream of which the subject is totally unaware, would be as incoherent as that of, e.g. a pain the subject does not feel. To a dream, as that noun is ordinarily understood, we could apply the Berkeleian tag, esse est percipi. Our awareness of a dream is both necessary and sufficient for its existence. On this view, dreams are conceived as mental ‘experiences’ that occur in sleep, private episodes in consciousness, which may or may not be remembered upon waking, and which may, if remembered at all, be quickly forgotten. In 1959, however, that conception was challenged in a strange but striking monograph by Norman Malcolm. In his view, the traditional notion of dreams as mental experiences was fundamentally misguided. Our concept of dreaming is derived not from any such experiences, but from the familiar practice of telling dreams when we wake up. According to Malcolm, it is a person's waking report of a dream, based upon a sincere memory claim, that is the sole criterion for the supposition that that person has dreamt. Thus Malcolm contended that 'in our daily discourse about dreams what we take as determining beyond question that a man dreamt is that in sincerity he should tell a dream or say he had one’ (59). Equally, the content of a dream report

must be the sole criterion for what that person has dreamt:

'His waking

impression is what establishes that he had a dream, and his account of his dream establishes what the content of his dream was' (79). Consequently, no sense could be attached to the idea that someone's memory of a dream was mistaken or incomplete, or had misrepresented the content of an original ‘experience’. There is, and could be, no evidence for the occurrence of any such experience. For the assumption that a person was aware of anything at all while asleep is ‘nonsensical’ (66). We may sometimes wonder whether we have had a dream that we fail to remember; or we may wonder whether the dream we seem to remember is the one that we really had.

But for Malcolm these are idle speculations, since

the questions they raise are unanswerable. There is no way we can revisit our sleeping experience to see if our memories of it are correct. Nor are there, in the nature of the case, any other sources to which we can appeal for confirmation of what we seem to remember having dreamt. Since there is no way of verifying our memories of dreams, it is futile to speculate about experiences whose occurrence there could be no reason either to affirm or to deny. A dream narrative is not, then, to be understood as a correct or

50

INTRODUCTION

mistaken description of the subject's experiences during sleep. Rather, if we seem to remember, upon waking, certain things that have not really been happening while we were asleep, we are properly said to have 'dreamt' those things. That is the role of such locutions as 'I dreamt that ...' in our language

(66). Malcolm's account has had no impact upon empirical psychology (where it remains largely unknown), and even among philosophers it has been widely rejected.* It would not be useful here, even if it were possible, to rehearse the debate it provoked. But it stimulated much valuable philosophical writing on dreams; and it will serve as a useful foil for viewing Aristotle, because its discussion of his text, even though partly mistaken, points to some troublesome questions at the ‘growing edge’ of his thought. If Malcolm were right, many questions traditionally asked about dreams should be abandoned as misconceived. We should not ask, for example, what makes us have experiences during sleep, but rather what makes us seem to remember having had such experiences. What would need explanation would not be phenomena occurring during sleep, but our waking impressions and verbal narratives based upon those impressions. A great deal of Aristotle's

theorizing,

if it assumes

the

traditional

framework,

would

therefore be misdirected. Aristotle certainly does embrace the traditional view. That view is embedded in the ordinary Greek word for dream, enupnion, which means literally ‘in sleep’, and in the common understanding of that word that Aristotle invokes (459a18-20); it reappears in his formal definition of a dream

(462a29—31,

cf. 462a16);

it is found in his causal account of dream

‘appearances’ (461a8-30, 461b67-462a8); and indeed it may be said to pervade his whole discussion from beginning to end. Near the start of the De Somno Aristotle asks whether it might be the case that people 'always dream when asleep, but do not remember’ (453b19). 'Always' here might mean (a) ‘every time they sleep’, or (b) ‘for the whole time they are asleep'. On the first interpretation, the possibility envisaged is that for at least part (not necessarily all) of each period of sleep, people have dreams which they do not remember. On the second, stronger (but much less likely) interpretation, it is envisaged that they dream throughout each period of sleep, yet do not remember all (but at most only some) of the dreams they

98

The most important critiques are anthologized by Dunlop, with the exception of that by Putnam, who demolished the 'verificationist' doctrines about meaning upon which Malcolm had implicitly relied.

INTRODUCTION

SI

have had. On either interpretation, Aristotle 15 asking whether we might, on at least some occasions, dream without subsequentiy remembering that we have dreamt or what we have dreamt. He does not, to be sure, endorse the theory that people ‘always dream’, even on the weaker interpretation. He nowhere returns to it explicitly, and it would presumably be incompatible with his own account of dreams. For under certain physical conditions specified in that account, no dreams will appear to the sleeper at all (461a11—21, 462b4-8, cf. Probl. 957a9—17). Nevertheless, Aristotle at least speaks as if the theory might conceivably be true, and would then itself call for explanation (453b20); he does not dismiss

it as the empty speculation that, on Malcolm's view, it would have to be. Moreover, even if Aristotle would have rejected the theory that 'we always dream', he must have supposed that we do sometimes fail to remember the dreams we have had. This is implied by his speaking of applying our minds and ‘trying to remember' (458b19, cf. 462a9-11). If it takes a conscious effort to remember something, that effort may be more or less successful, or may fail altogether. The experience of trying to remember a dream, but being aware of partial or complete failure, is common enough, and it is not clear what account should be given of it on Malcolm's view.

Aristotle does not, like Malcolm,

treat our waking memories as criteria for our having dreamt, or for the content of our dreams. He simply appeals to waking memory to support a distinction between dream appearances and thoughts occurring in sleep. His argument requires the assumption that waking memory is generally reliable as a guide to the nature of our experience while asleep. But unlike Malcolm's, it does not require that our memories of that experience must, on

conceptual grounds, be incapable of error. Aristotle makes various other remarks that would be hard to accommodate from Malcolm's perspective. For example, he contends that certain animals dream (463b12). This must, on Malcolm's view, appear an empty hypothesis, since in their case the criteria of 'reports' or "waking memories' could not possibly be satisfied. Aristotle remarks (HA 536b30) that dogs show that they are dreaming by whining in sleep. Their whines do not, of course, constitute dream ‘reports’, as Malcolm recognizes: "We say of a dog, when he whines and twitches his feet in sleep, "He must be having a dream": and here there 1s no question of what he will tell us when he wakes up' (62).

But, he

counters, 'this use of language is not quite serious: one draws no practical consequences from the supposition that a dog is dreaming'. He also says that

52

INTRODUCTION

whereas a human dream has a content, which is described when it is related, a dog's dream has none (63). Unfortunately, it remains unclear why Malcolm thinks that our use of language in ascribing a dream to a sleeping dog is ‘not quite serious’, or why it should be made so by our drawing no ‘practical consequences’ from the ascription. What ‘practical consequences’ must we draw from the supposition that human beings are dreaming, to ensure that our ascriptions of dreams to them are a fully 'serious' use of language? And why may we not infer the content

of a dog's

dreams

from

his whines,

twitches

or other behaviour

during sleep? Perfectly serious attempts have been made to do so? Verbal language doubtless gives human dreams incomparably richer content than any animal dream could possess. But in this respect dreams are in no way peculiar. Similar claims could be made for many other human experiences vis-à-vis those of animals. The case of animals suggests that we may properly use behavioural criteria for the attribution of dreams, and need not depend entirely upon verbal narratives. On this point Aristotle's position is intuitively far more plausible than Malcolm's. A more

central issue, however,

is raised by Aristotle's contention

that

during sleep we may be aware that we are asleep and dreaming: ‘for often something in the soul of a person asleep says that what is appearing is a dream’ (462a5-7, cf. 45946--8). He also notes that sleeping people may be © dimly aware of external phenomena, such as the light of a lamp or the sounds of animals, and may sometimes even answer questions, presumably (though Aristotle does not say so) about such things, put to them while they are still asleep (462a25-26). Elsewhere (GA 779a16-19, cf. DS 456a24-29) he mentions somnambulism. Sleepers may get up and do many things without dreaming: they can walk about 'and see as well as anyone awake’, being aware through their senses of what is going on. All of these passages state or imply that various kinds of awareness during sleep are possible. Malcolm quotes without dissent (29—30) the passage about people who have faint sense-impressions during sleep (462a19-27). But he thinks that our ascription of sleep to such people must be qualified. For in those cases the criteria for being sound asleep cannot be fully satisfied. He defines 'sound asleep' in such a way that people in that state must be unable to assert or to judge that they are in it. They must, in this respect, be like those who are unconscious or dead. For if they should assert or judge 99

Anextended and plausible example in antiquity is provided by Lucretius, IV. For a modern discussion, see Evans, 208-11.

987-1010.

INTRODUCTION

53

that they are in one of those states, the conditions necessary for their making that assertion or judgment must falsify its content: if it were true, they could not be making it; hence, if they are making it, it cannot be true (6-7). Malcolm then extends and generalizes his argument: we can have no evidence that any assertion or judgment is ever made by anyone during a state of ‘sound sleep’. Accordingly (3-7), Aristotle could have had no basis for saying, in his opening example of a dream (458b10-11), that during sleep ‘we say not only that the thing approaching is a man or a horse, but also that it is pale or beautiful'. It must, then, be meaningless either to assert or to deny that during sound sleep ‘we say' any such things. Later the argument is extended from assertion and judgment to mental phenomena quite generally. Thinking, reasoning, having imagery, feeling emotions, having illusions or hallucinations,

during

a

state

of

'sound

sleep'

are

all,

for

Malcolm,

‘unintelligible notions' (45). Consequently, dreaming cannot consist in the occurrence of any such mental phenomena. Aristotle's opening example of the man or horse dream deserves to be considered in its context. He 15 arguing at that point (458b10—25) that dreaming is not an ordinary exercise of judgment. When, in a dream, we seem to see a pale man or a beautiful horse, we cannot, it would appear, be judging that something before us is a man or a horse, is pale or beautiful. For we could not make such judgments, truly or falsely, without concurrent perception (458b11—13); yet we are not, while asleep with our eyes shut, perceiving anything, as has just been argued (458b7-9). So dreaming of a pale man or a beautiful horse cannot, in the ordinary way, be judging (even falsely) that something perceived is a man or a horse, is pale or beautiful, since ex hypothesi during sleep nothing is being perceived at all. This argument is dialectical: Aristotle is, at this early stage, only setting up the problem of dreaming, not yet solving it. Ironically, his argument, so far as it goes, actually supports the Malcolmian view that dreaming should not be identified with making a judgment during sleep. But this is not because Aristotle views the notion of making a judgment during sleep as ‘unintelligible’. It is simply because (for reasons discussed above)! he draws a sharper distinction than we should be inclined to allow between dream appearances and thoughts entertained in sleep. Since he evidently thinks that both dream appearances and judgments can, in fact, occur simultaneously during sleep, it is clear that he remains wedded to the view that Malcolm attacked. 100

See pp. 6-10.

54

INTRODUCTION

It is not at all obvious who has the better of the issue between them. Aristotle's contention (462a2~7) that we sometimes become aware that we are asleep and dreaming seems persuasive. On the other hand, it is usually with the dawning of such awareness that our dreams end abruptly, and we find ourselves awake. We may therefore prefer not to say that we become aware of being asleep or of dreaming, but rather that we become aware of having been asleep, and of having dreamt. Thus, we may share Malcolm's discomfort in holding that we can be aware of being asleep and of dreaming, while we are still sound asleep. This discomfort may be extended to any kind of awareness during sleep, if sleep is defined as a state of perceptual incapacitation. Aristotle must, as we noticed earlier, face a problem in explaining how any perceptual awareness, including awareness of dreams, can occur in sleep, if that is defined in the terms of the De Somno as a state of general perceptual shutdown. He himself betrays discomfort relevant to this problem, when he remarks at 462a26—27, that ‘in the case of sleeping and waking, when one of these states is present in the ordinary way, it is possible for the other to be present in a certain manner’. For this remark seems to anticipate just the sort of objection that a Malcolmian critic might have put to Aristotle: ‘how can people perceive anything, even faintly, and how can they understand and answer questions put to them, if they are sound asleep? Must not all those things be ruled out, given the terms in which you have defined and explained sleep?’ Analogous questions might be asked regarding Aristotle's account of dreams. How can we be 'conscious' of our dreams, in the way he suggests, while we are completely 'unconscious' in virtue of being sound asleep? Would Aristotle be content to answer that while dreaming we are both 'asleep in the ordinary way’, yet also ‘in some sense awake'? Presumably that answer would not satisfy him, given his distinction between dreaming and faint awareness of sense-impressions originating externally. Since sleep is insistently built into his definition of dreaming (459a19-20, 462a30-31), it is doubtful that he would allow that a dreamer is ‘in some sense awake’. At all events, he nowhere suggests using that answer in connection with dreams. But then his problem persists. Perhaps he starts to grapple with it in the first chapter of the De Insomniis, as we have earlier suggested,’ but it is not clear that he wrestles it to the ground. 101

Pp. 25,28.

102

Pp. 20-1 above.

INTRODUCTION

55

Something more can be learnt from a comparison between Aristotle and Malcolm. For even if Malcolm's conclusions are unacceptable, his work reminds us of an important aspect of dreaming that Aristotle's discussion may lead us to overlook. A genetic account of dreams, such as Aristotle's, will not suffice for understanding dreams philosophically. For it will not capture the relation between a dream and its content, between the dream and what it is a

dream 'of. What makes my dream of Coriscus a dream of him? relation, comparable

This is not a causal

with the relation between a reflection and that which

casts it. What makes a dream to be one of Coriscus is not the fact that an appearance in sleep is caused by a prior waking perception of him. It could be that we very often do have nocturnal presentations of individuals we have previously seen, and those presentations could be caused by just such a physical mechanism as Aristotle describes. But, even if all that were granted, it would not be simply in virtue of that causal connection that we would be said to have dreams of those individuals. Here it is important to see that the trouble does not lie in the details of Aristotle's causal story. We could amend that story to take account of the most

advanced

neurological

research,

but

we

still

would

not

have

a

satisfactory conceptual understanding of what it is to have a dream of Coriscus. This will become clear if we reflect that we may dream of individuals for which no such causal story would be possible, e.g. of historical or mythical characters, such as Napoleon or Odysseus. For no waking perception of such long dead or mythical individuals could be said to have caused the dream. We might, to be sure, suggest some genetic explanation of dreams of such figures. But even more clearly than in dreams of living people, the causal story would fail to explicate the relation between these dreams and their objects. It is easy to be misled here by Aristotle's comparison of dreams with reflections (461a14-25, 464b5-16). For the relation between an object and its reflection is a causal one. If we see a reflection cast by a man, so badly distorted that we mistake it for a reflection of a horse, then our mistake can be corrected. We need only be shown the causal relation holding between the man and the reflection, and that settles the matter:

it is a reflection of a man,

not of a horse. But having a dream of Coriscus is not like seeing a reflection of him. It is not possible to mistake a dream of one person for a dream of another. Even if the figure in our dream were to look very unlike Coriscus but very like (say) Socrates, and even if it could be traced causally to our prior perception of Socrates, it would still be a dream of Coriscus, provided it

56

INTRODUCTION

was Coriscus whom we pronounced the dream figure to be. For in our dream Coriscus might have been strangely endowed with the features of Socrates. That might have been how we represented him in a weird flight of fancy. The relation between

Coriscus

and our dream of him is, in this way, more

like the relation between Coriscus and a picture we are painting of him, or between Coriscus and a story we are making up about him. For here too we may represent him according to our fancy, and not necessarily as he really is. The identity of the dream figure depends upon the dreamer's say-so, somewhat as the identity of a picture's object depends, at least in part, upon what the artist intends or claims to depict. Malcolm was therefore right to hold that a dreamer's own story is, in some way, crucial for the determination of what has been dreamt. Dreamers' accounts of what their dreams are ‘of do have a special status, comparable with painters’ accounts of what their pictures are 'of', or with story-tellers' accounts of what their tales are about. But it does not follow that dreamers' reports must 'establish' the content of their dreams beyond question, in the sense that the report must necessarily be immune from all possibility of error. For the story told in a waking report may well differ from the one originally 'told' by the dreamer while asleep. Our dreams may, in some respects, 5 be usefully viewed as 'stories' that we tell ourselves while we are asleep. Such stories may be re-told in a report upon waking up. Our later reporting of a dream is, we might say, a narration of an earlier narration, a recounting of our own previous fiction-telling. The original story is as immune from error as is any work of fiction (hence we do not speak of dreams as being either 'correct' or mistaken). But our waking report is as capable as any other report of being mistaken. For upon waking we may, wittingly or unwittingly, misrepresent the story that we 'told' ourselves in sleep. Or we may often, perhaps nearly always, forget the story altogether.

If that view of dreaming

is acceptable, we might mediate between

Aristotle and Malcolm in the following way. Malcolm was right to give dreamers' reports a special status in the determination of what they have dreamt. But he was wrong to suggest that waking impressions or reports of dreams could not be mistaken, or that for someone to be properly said to have dreamt, nothing need have happened in sleep at all. Aristotle was justified in assuming that dreams occur during sleep. But his genetic account was 103

Not in all. For, unless we posit an unconscious agency, purposive explanation. See sections 7 and 10 above.

dreams

do not admit of

INTRODUCTION

57

insufficient to explain the relation between dreams and their objects. To treat dreams as products of the imagination was a correct insight on his part. But the implications of that insight for a conceptual understanding of dreaming still remain to be fully worked out.

12.

Conclusion

This essay has more often sought to understand Aristotle than to defend him. His accounts of sleep and dreams raise major problems of interpretation and leave many questions unanswered. Other questions he could not even have

asked. The scope of his inquiry was narrowly limited, and much information relevant to it lay beyond his reach. But none of those points diminishes the significance of his work. He inherited traditional beliefs about dreams that it was no small achievement to have undermined. Nor were these matters to be any better understood for centuries to come. Until recently one might have shared the doubt voiced by Dodds (120) in 1949 whether modern science had advanced very far beyond him. Half a century later, Dodds's judgment sounds outdated. It preceded by four years the discovery of rapid eye-movement sleep, and the vast proliferation of work in experimental psychology that has ensued. Today the whole field of sleep and dream research is once more in ferment. But much of the ferment is in the spirit of Aristotle's own endeavours. We can be sure that he would approve of the path that scientific inquiry into sleep and dreams is now following. For on that path, as on so many others, he was the greatest of ancient pioneers.

Text and Translation

60

453b

ΠΕΡῚ

KAI

YIINOY

ἘΕἘΓΡΗΓΟΡΣΕΩΣ

1. Περὶ δὲ ὕπνου καὶ ἐγρηγόρσεως ἐπισκεπτέον τίνα τε τυγχάνει ὄντα, καὶ πότερον ἴδια τῆς ψυχῆς ἢ τοῦ σώματος ἢ κοινά, καὶ εἰ κοινά, τίνος μορίου τῆς ψυχῆς ἢ τοῦ σώματος, καὶ διὰ τίν’ αἰτίαν ὑπάρχει τοῖς ζῴοις: Kai | πότερον ἅπαντα κεκοινώνηκεν ἀμφοτέρων, ἢ τὰ μὲν θατέρου τὰ δὲ θατέρου μόνον, ἢ τὰ μὲν οὐδετέρου τὰ δὲ ἀμφοτέρων: πρὸς δὲ τούτοις τί ἐστι τὸ ἐνύπνιον, καὶ διὰ τίνα αἰτίαν οἱ καθεύδοντες ὁτὲ μὲν ὀνειρώττουσιν ὁτὲ δὲ οὔ, ἢ συμβαίνει μὲν ἀεὶ τοῖς καθεύδουσιν ἐνυπνιάζειν, ἀλλ᾽ οὐ μνημονεύουσιν, | καὶ εἰ τοῦτο γίγνεται, διὰ μέλλοντα τὰ ἐνδέχεται καὶ πότερον τίνα αἰτίαν γίγνεται: προορᾶν ἢ οὐκ ἐνδέχεται, καὶ τίνα τρόπον εἰ ἐνδέχεται: καὶ πότερον Τὰ μέλλοντα ὑπ᾽ ἀνθρώπου πράσσεσθαι μόνον, ἢ καὶ ὧν ἢ ἀπὸ γίγνεται ἔχει τὴν αἰτίαν, καὶ φύσει τὸ δαιμόνιον ταὐτομάτου.

πρῶτον

μὲν

οὖν

| τοῦτό

γε

φανερόν,

ὅτι

τῷ

αὐτῷ

τοῦ

ζῴου

15

20

25

ἥ τε ἐγρήγορσις ὑπάρχει καὶ ὁ ὕπνος: ἀντίκεινται γάρ, καὶ φαίνεται στέρησίς τις ὁ ὕπνος τῆς ἐγρηγόρσεως: ἀεὶ γὰρ τὰ ἐναντία καὶ ἐπὶ τῶν ἄλλων καὶ ἐν τοῖς φυσικοῖς ἐν τῷ αὐτῷ δεκτικῷ φαίνεται γιγνόμενα, καὶ τοῦ αὐτοῦ ὄντα πάθη, λέγω δ᾽

οἷον

ὑγίεια

ἀσθένεια,

| νόσος,

ὄψις

καὶ

δὲ

καὶ

ἐκ

γνωρίζομεν,

τούτῳ

ἔτι

καὶ

καὶ

καὶ

καὶ

κάλλος

τυφλότης,

τῶνδε τὸν

καὶ

δῆλον:

καὶ

αἶσχος,

ἀκοὴ

καὶ



γὰρ

καθυπνοῦντα:

τὸν

καὶ

ἰσχὺς

καὶ

30

κωφότης.

τὸν δὲ

ἐγρηγορότα

4548

αἰσθανόμενον

ἐγρηγορέναι νομίζομεν, καὶ τὸν ἐγρηγορότα πάντα ἢ τῶν ἔξωθέν τινος αἰσθάνεσθαι ἢ τῶν ἐν αὑτῷ κινήσεων. εἰ τοίνυν τὸ ἐγρηγορέναι ἐν μηδενὶ ἄλλῳ ἐστὶν ἢ τῷ αἰσθάνεσθαι, δῆλον ὅτι QUep αἰσθάνεται, τούτῳ καὶ ἐγρήγορε τὰ ἐγρηγορότα καὶ

5

4530

61 ON SLEEP AND WAKING Chapter One 1. 453b11-24. Survey of questions regarding (1) sleep and waking; (2) dreams; and (3) precognition based upon dreams.

15

453b11. As for sleep and waking, we must consider what they are, and whether they are proper to the soul or to the body, or are common to both. If they are common, which part of the soul and the body has them, and from what cause do they belong to animals? And | do all animals partake of both, or do some partake of one only, and some of the other? Or do some partake of neither, and some of both?

20

In addition, we must ask what

a dream is, and from what cause sleepers sometimes dream but sometimes not. Or is it the case that people always dream when asleep, but do not remember? i And from what cause does this happen, if it does? Again, is it or is it not possible to have prevision of things that are going to happen? And if this is possible, in what way? Is it possible only with actions that are going to be performed by man? Or does it also cover things for which the daemonic is responsible, i.e. things that happen by nature or fortuitously? 2. 453b24—454a11. Waking and sleep are proper neither to the soul nor to the body, but both belong to the same part of an animal, i.e. the part whereby it perceives. For (1) they are opposites, and must therefore inhere in the same subject; and (2) they are recognized by the presence or absence of a common criterion, namely, perception.

25

30

453b24. First, then, | this much is plain: waking and sleep both belong to the same part of an animal. For they are opposed to each other, and sleep is evidently a sort of privation of waking. For in natural things, as in other cases, opposites evidently come into being in the same recipient, and are affections of the same subject. I mean for example health and | sickness, beauty and ugliness, strength and weakness, sight and blindness, hearing and deafness.

454a 453b31. Again, | the point is clear from the following. We recognize a person as sleeping by the same mark as that by which we recognize someone as waking. It is the person who is perceiving that we consider to 5 be awake; and we take every waking person to be | perceiving either something external or some movement within himself. If, then, the waking state consists in nothing else but perceiving, it is clear that waking

62

ΠΕΡῚ

YHNOY

KAI

ΕΓΡΗΓΟΡΣΕῺΩΣ

4548

καθεύδει τὰ καθεύδοντα. ἐπεὶ δὲ οὔτε τῆς ψυχῆς ἴδιον τὸ αἰσθάνεσθαι οὔτε τοῦ σώματος (οὗ γὰρ ἡ δύναμις, τούτου καὶ ἡ ἐνέργεια: ἡ δὲ λεγομένη αἴσθησις ὡς ἐνέργεια κίνησίς τις διὰ τοῦ σώματος τῆς ψυχῆς ἐστι), φανερὸν ὡς οὔτε τῆς ψυχῆς τὸ πάθος ἴδιον, οὔτ᾽ ἄψυχον σῶμα δυνατὸν αἰσθάνεσθαι.

10

διωρισμένων δὲ πρότερον ἐν ἑτέροις περὶ τῶν λεγομένων ὡς μορίων τῆς ψυχῆς, καὶ τοῦ μὲν θρεπτικοῦ χωριζομένου τῶν ἄλλων ἐν τοῖς ἔχουσι σώμασι ζωήν, τῶν δ᾽ ἄλλων οὐδενὸς ἄνευ τούτου, [ δῆλον ὡς ὅσα μὲν αὐξήσεως καὶ φθίσεως μετέχει μόνον τῶν ζώντων, τούτοις οὐχ ὑπάρχει ὕπνος οὐδὲ ἐγρήγορσις, οἷον τοῖς φυτοῖς (οὐ γὰρ ἔχουσι τὸ αἰσθητικὸν μόριον, οὔτε εἰ χωριστόν ἐστιν οὔτε εἰ μὴ χωριστόν: τῇ γὰρ δυνάμει καὶ τῷ

εἶναι

χωριστόν ὁμοίως ft

καθεύδει,

καὶ

x

ἀλλὰ

»

ἐστιν).

δὲ

LA

2

15

ὅτι

Ν

οὐδέν

t,

τοῖς

x

2

αὐτοῖς

a

+

ἐστιν

LA

t



*

ὑπάρχει

~

ἀεὶ

e

*

τῶν

iA

ἐγρήγορεν

^

*

ζῴων

^



/,

ἀμφότερα

ra

3

ἀεὶ

nn

τὰ

,

3

20

^

πάθη

\

[A

ταῦτα. οὐ γάρ, εἴ τι ἔστι ζῷον «μὴ» ἔχον αἴσθησιν, τοῦτ᾽ ἐνδέχεται οὔτε καθεύδειν οὔτε ἐγρηγορέναι: ἄμφω γάρ ἐστι Ta πάθη ταῦτα περὶ αἴσθησιν τοῦ πρώτου αἰσθητικοῦ. *

,



οὐκ

L4

ἐνδέχεται

+

bi

δὲ

΄

*

3^

οὐδὲ ,

»

L4

θάτερον

bi

LA

τούτων

5

^

ἀεὶ

#

3

,

x

ὑπάρχειν

^N

f

| TH

^

^

αὐτῷ, olov ἀεί Ti γένος ζῴων καθεύδειν ἢ ἀεί τι ἐγρηγορέναι, ἐπεὶ ὅσων ἔστι τι ἔργον κατὰ φύσιν, ὅταν ὑπερβάλλῃ τὸν χρόνον ὅσον δύναταί τι ποιεῖν, ἀνάγκη ἀδυνατεῖν, οἷον τὰ ὄμματα ὁρῶντα, καὶ παύεσθαι τοῦτο ποιοῦντα, ὁμοίως δὲ καὶ χεῖρα καὶ ἄλλο πᾶν οὗ ἔστι τι ἔργον. ei δή τινός ἐστιν ἔργον τὸ A

*

^

t^

v

we

^

3

LA

+

^

^

v

^

καὶ

αἰσθάνεσθαι

συνεχῶς,

2.

*

%

αἰσθάνεσθαι,

1.

25

T

τοῦτο,

»

ἂν

,

t^

3

t

LA

΄

ὑπερβάλλῃ

ἀδυνατήσει

καὶ

tU

ὅσον



FA

,ὔ

3

^

ἣν

οὐκέτι

*

la

Y



χρόνον

τοῦτο

x

4

δυνάμενον

ποιήσει.

εἰ

The translation is based upon Ross's emended text at 454a21, which adds μὴ before ἔχον. Similarly Beare ad loc. Alternatively, omitting μὴ, ‘For no animal which has perception can be neither asleep nor awake’ (Hett). But see Beare, and Ross (255-6). Siwek (197, n. 18) defends the unemended text. The translation at 454226, ‘For all things’, assumes a comma after ἐγρηγορέναι, and requires either ὅτι (Becker) or ἐπεὶ (Susemihl) for the mss' ἔτι, printed by Ross.

Thus also Beare and Hett.

Ross's text would require the translation, ‘Moreover, all

things ..'. But 'Moreover would mark the sentence as starting a new argument for the conclusion that no kind of animal could be always awake. Yet no argument at all for that conclusion has yet been given. The variant reading εἰ for ἔτι in one ms (P) would give the required linkage with b24—25, with ‘if instead of ‘for’ or ‘because’. But this variant also gives very strong plausibility to ἐπεὶ, which is rightly preferred by Barnes.

30

4548

10

ON SLEEP

63

things are awake, and sleeping things are asleep, with the same part as that whereby they perceive. But given that perceiving belongs neither to the soul nor to the body solely (for what owns any capacity also owns its exercise; and what is called perception, in the sense of exercise, is a certain movement of the soul by means of the | body), it is plain that the affection is not peculiar to the soul, nor is a soul-less body capable of perceiving. 3. 454a11-454b9. Neither sleep nor waking can characterize plants. Both States must characterize all animals. No animal could be always awake.

15

454a11. Now we have already distinguished elsewhere what are called ‘parts’ of the soul. The nutrient part in bodies possessing life can exist separately from the others, whereas none of them can do so without it. Accordingly, | it is clear that neither sleep nor waking belongs to living things that partake only of growth and decay, such as plants. For they do not possess the perceptual part, whether or not it be capable of separate existence. For both as a capacity and in its being, it is separable.

20

454219. It is likewise clear that there is nothing that is always awake | or always asleep. Rather, both these affections belong to the same animals. For if there is any animal that does not possess perception,! it is not possible that it should either sleep or wake. For both these affections have to do with perception exercised by the primary perceptual part.

25

454a24. Nor is it possible that one of them should always belong to | the same animal, e.g. that a certain kind of animal should be always asleep or always awake. For? all things that have a natural function must, whenever they exceed the time for which they can do a certain thing, lose their capacity and cease from doing it, e.g. the eyes from seeing. Likewise the hand, and everything else that has a function. So if | perceiving is the function of some part, then this part too, should it exceed the due time for which it is capable of perceiving continuously, will lose its capacity and

30

64

ΠΕΡῚ

YIINOY

KAI

ἘΓΡΗΓΟΡΣΕΩΣ

454a

τοίνυν τὸ ἐγρηγορέναι ὥρισται τῷ λελύσθαι τὴν αἴσθησιν, τῶν δ᾽ ἐναντίων τῶν μὲν ἀνάγκη θάτερον ἀεὶ παρεῖναι τῶν δ᾽ οὔ, τῷ δ᾽ ἐγρηγορέναι τὸ καθεύδειν ἐναντίον, καὶ ἀναγκαῖον ἅπαντι θάτερον ὑπάρχειν, ἀναγκαῖον ἂν εἴη καθεύδειν. εἰ οὖν τὸ τοιοῦτον πάθος ὕπνος, τοῦτο δ᾽ ἐστὶν ἀδυναϊίμία 8v ὑπερβολὴν τοῦ ἐγρηγορέναι, ἡ δὲ τοῦ ἐγρηγορέναι ὑπερβολὴ ὁτὲ μὲν νοσώδης ὁτὲ δὲ ἄνευ νόσου γίγνεται, ὥστε καὶ ἡ ἀδυναμία καὶ ἡ διάλυσις ὡσαύτως ἔσται, ἀνάγκη πᾶν τὸ ἐγρηγορὸς ἐνδέχεσθαι καθεύδειν: ἀδύνατον γὰρ ἀεὶ ἐνεργεῖν. ὁμοίως δὲ οὐδὲ καθεύδειν οὐδὲν ἀεὶ ἐνδέχεται. ὁ γὰρ ὕπνος πάθος τι τοῦ αἰσθητικοῦ μορίου ἐστίν, οἷον δεσμός τις καὶ ἀκινησία, ὥστ᾽ ἀνάγκη πᾶν τὸ καθεῦδον ἔχειν τὸ αἰσθητικὸν μόριον. αἰσθητικὸν δὲ τὸ δυνατὸν αἰσθάνεσθαι Kat’ ἐνέργειαν: ἐνεργεῖν δὲ τῇ αἰσθήσει κυρίως καὶ ἁπλῶς ἀδύνατον καθεῦδον ἅμα: διὸ ἀναγκαῖον ὕπνον πάντα ἐγερτὸν εἶναι. x

A

"

^

^

,

τὰ μὲν | οὖν ἄλλα σχεδὸν ἅπαντα δῆλα κοινωνοῦνθ᾽ ὕπνου, καὶ πλωτὰ καὶ πτηνὰ καὶ πεζά: καὶ γὰρ τὰ τῶν ἰχθύων γένη πάντα καὶ τὰ τῶν μαλακίων ὦπται καθεύδοντα, καὶ τἄλλα πάντα t

v

ὁσαπερ

ἔχει

,

#

ὀφθαλμούς’

καὶ

Ν

x

γὰρ

x

τὰ

7

σκληρόφθαλμα

φανερὰ

x

^

τοῦτο

3.

4.

εἰ

,

καθεύδουσιν. #

εἰ

*

δέ

F4

Ty

πιθανὸς x



f

λεχθεὶς x

10

15

x

καὶ

τὰ ἔντομα κοιμώμενα- βραχύυπνα δὲ τὰ τοιαῦτα | πάντα, διὸ καὶ λάθοι ἄν τινα πολλάκις πότερον μετέχουσι τοῦ καθεύδειν ἢ οὔ. τῶν δ᾽ ὀστρακοδέρμων κατὰ μὲν τὴν αἴσθησιν οὐδέ πω γέγονε φανερὸν

454b

λόγος, ,F

πεισθήσεται.

The translation follows Ross's text at 454b1—2, based on the reading of ΕΜ. Hett reads with LSU τὸ μὲν ἀνάγκη παρεῖναι τὸ 8° où, translating, ‘if of two contraries one must be present and the other not’. But the contrast required is not between two members of a single pair of opposites. The generalization in Hett's version does not hold true for all opposites, since it does not hold for 'contraries', which were distinguished from 'contradictories' by Aristotle himself (Categories 1221-25, De Interpretatione 17b16-25). The required contrast is rather between those pairs of opposites for which the generalization does hold true and those for which it does not. Aristotle's argument is that if sleeping and waking are opposites of the latter kind, then sleeping must be necessary. Thus rightly Ross (256) and Beare ad loc. Reading τοῦτο at 454b23, with Bywater, Beare, Hett, and Ross. Keeping the mss' τούτῳ, we must translate, 'If anyone has found the foregoing argument persuasive, he will be persuaded by this‘. But since 'this' can refer only to ‘the foregoing argument’, the mss' text makes the sentence tautologous.

20

454a

ON SLEEP

65

454b will do so no longer. If, therefore, being awake is defined by the | freeing of perception; and if for some opposites it must always be true that one of the pair is present, while for others this is not so? and if sleeping is opposite to waking, and one of this pair must belong to every animal; then sleeping would have to be necessary. If, then, such an affection is sleep, 5 and this is an incapacitation | due to excess of waking; and if an excess of waking is sometimes unhealthy but sometimes occurs without illness, so that both

the incapacitation

and

the dissolution

will be similar;

then,

necessarily, everything that wakes must be capable of sleeping. For it is unable to be active at all times. 4. 454b9-455a3. Nor could anything be always asleep. Sleep is observable in almost every animal species, including fish and insects. It must be universal in animals, by contrast with plants. 10

15.

20

454b9. It is likewise impossible for anything to be always sleeping. For | sleep is a certain affection, a sort of 'fettering' or immobilization of the perceptual part. Consequently, everything that sleeps must be possessed of the perceptual part. But it is what can actively exercise perception that is perceptual. And it is impossible for anything to be actively exercising perception in the chief and unqualified sense, while sleeping at the same time. Hence all sleep must be subject to awakening.

454b14. Now | almost all other animals are observed to partake of sleep, aquatic, winged, and terrestrial creatures alike. For every kind of fish and the soft-shelled species have been seen sleeping, as has every other creature that has eyes. For it is plain that even hard-eyed creatures and insects take repose, although all such things have short spells of sleep, | so that it could often escape one whether they partake of sleep or not. For testaceans, direct observation has not yet made it plain whether they do sleep. But anyone who has found the foregoing argument convincing will

be persuaded of this.4

ΠΕΡῚ

ὅτι μὲν οὖν τούτων. τῷ γὰρ αἰσθήσεως τρόπον ὕπνον εἶναί φαμεν, δὲ φυτῶν οὐδὲν παθημάτων: ἄνευ οὔτε ἐγρήγορσις"

ὕπνου κοινωνεῖ τὰ ζῷα πάντα, φανερὸν ἐκ αἴσθησιν | ἔχειν ὥρισται τὸ ζῷον, τῆς 8 τινὰ τὴν μὲν ἀκινησίαν καὶ οἷον δεσμὸν τὸν τὴν δὲ λύσιν καὶ τὴν ἄνεσιν ἐγρήγορσιν. τῶν οἷόν τε κοινωνεῖν οὐδετέρου τούτων τῶν μὲν γὰρ αἰσθήσεως οὐχ ὑπάρχει οὔτε ὕπνος οἷς δ᾽ αἴσθησις ὑπάρχει, καὶ τὸ λυπεῖσθαι καὶ

x

΄

τὸ

Ν

T

χαίρειν:

^

δὲ

oig

YIINOY

KAI

454b

66

N

ταῦτα,

ἘΓΡΗΓΟΡΣΕΩΣ

1

,

ἐπιθυμία’

καὶ

x

^

m

δὲ

τοῖς

+

φυτοῖς

Y

οὐδὲν

ὑπάρχει τούτων. σημεῖον δ᾽ ὅτι καὶ τὸ ἔργον τὸ αὑτοῦ ποιεῖ τὸ θρεπτικὸν μόριον ἐν τῷ καθεύδειν μᾶλλον ἢ ἐν τῷ ἐγρηγορέναι: τρέφεται #

πρὸς

γὰρ

καὶ

\

ταῦτα

αὐξάνεται

M

τότε

[A

,

f

μᾶλλον, ^

ὡς

οὐδὲν 5



25

455a

προσδεόμενα

*

7

αἰσθήσεως.

τῆς

2. Διὰ τί δὲ καθεύδει καὶ ἐγρήγορε καὶ διὰ ποίαν τινὰ | αἴσθησιν ἢ ποίας, εἰ διὰ πλείους, σκεπτέον. ἐπεὶ δ᾽ ἔνια μὲν τῶν ζῴων ἔχει τὰς αἰσθήσεις πάσας, ἔνια δ᾽ οὐκ ἔχει, οἷον Y

b

ὄψιν,

τὴν

ζῴων 3

4

δ᾽

t

ἀτελές

΄

ἀδύνατον

δ᾽

x

x

ἀφὴν

καὶ

x

(εἴρηται

5

2

\

ἐστὶν

A

τὴν

t



γεῦσιν

δὲ

περὶ

^



ἁπλῶς

Y

ἅπαντα

αὐτῶν A

ἐν

3

δέ

Ei

^

ἀκούειν, δέ τις

5.

ΜΕ

t

΄

καθ᾽ Ν μὲν

+

£t

΄

Y

^

τῶν

ψυχῆς),

4

x

| τὸ

x

7

κατὰ τὸν αὐτὸν ἀκολουθοῦσα ^

Y 5

ἴδιον, > E ἀκοῇ

*

τὸ x TO

τρόπον, ἔστι πάσαις, À

T

Alternatively, reading πᾶσιν at 455a10 with several mss, 'the same affection must belong to all [sc.

animals].

Thus Hett.

Cf.

10

ὑπάρχειν τὸ αὐτὸ τῇ δὲ μή, ταύτῃ

ἑκάστην αἴσθησιν τὸ μέν τι < ^ Y b t ^ ^ οἷον τῇ ὄψει τὸ ὁρᾶν, τῇ δ᾽

καὶ ταῖς ἄλλαις ἑκάστῃ καὶ κοινὴ δύναμις

τι

αἰσθάνεσθαι

καθεῦδον ζῷον, φανερὸν ὅτι πάσαις ἀναγκαῖον πάθος ἐν τῷ καλουμένῳ ὕπνῳ εἰ γὰρ τῇ μέν, καθεῦδον αἰσθήσεται, τοῦτο δ᾽ ἀδύνατον. ἐπεὶ δ᾽ ὑπάρχει , Y c. τι κοινόν, ἴδιον

εἴ

περὶ

1

αἴσθησιν

Y

πλὴν

τοῖς

Y

ὁποιανοῦν

s

ἔχει,

also the conclusion drawn at 455a26—

27. But the hypothesis to be rejected at a11—12 is that one sense is affected, but another not (τῇ μὲν, τῇ δ᾽). The rejection of this hypothesis warrants the conclusion that the same affection belongs to all of the senses.

15

ON

454b 25

defined

30

455a

SLEEP

67

454b23. That all animals partake of sleep, then, is plain from these considerations. It is by its possession of perception | that an animal is as such.

And

we

maintain

that sleep

is, in a certain

way,

an

immobilization or 'fettering' of perception; whereas its liberation or release is waking. By contrast, no plant can partake of either of these affections. For neither sleep nor waking belongs to anything without perception; whereas things to which | perception belongs can also have pain and pleasure. And what can have these can also have appetite. But none of these belongs to plants. Proof of this is the fact that the nutrient | part performs its function more during the sleeping than the waking state. For more nutrition and growth take place at that time, suggesting that nothing is needed from perception to further those ends.

Chapter Two 5. 455a4-12. To which sense-faculty are sleep and waking due? All of the special senses must be similarly affected in sleep. For it is impossible to be perceiving with one sense during sleep but not with another. 455a4.We

10

must now consider why it is that one sleeps or wakes, and to

what sort of | sense (or what sorts, 1f more than one) those states are due. Now, given that some animals possess all of the senses, while others do not possess, e.g., sight, yet all possess touch and taste — apart from any animal that is imperfect (and these have been discussed in our work on the soul); given also that an animal, | while sleeping, cannot be experiencing any perception at all in an unqualified way; it is plain that during the state called sleep the same affection must belong to all of the senses? For if it belongs to one sense but not to another, then an animal will perceive with the latter while sleeping. But that is impossible. 6. 455a12—b2. There must be a capacity common to all of the special senses, a single sense-faculty with its own organ. This will serve as a master sense-organ, upon whose functioning that of the special senses depends.

15

455a12. Now also something hearing to that all are attended

to every sense there belongs something special to it, and common. Thus, seeing is special to the sense of sight, and of hearing, | and so on for the other senses severally. But also by a certain common faculty, whereby one perceives

68

ΠΕΡῚ

ΥΠΝΟΥ

KAI

455a

ἘΓΡΗΓΟΡΣΕΩΣ

καὶ ὅτι ὁρᾷ καὶ ἀκούει αἰσθάνεται: οὐ γὰρ δὴ τῇ γε ὄψει ὁρᾷ ὅτι ὁρᾷ, καὶ κρίνει δὴ καὶ δύναται κρίνειν ὅτι ἕτερα τὰ γλυκέα τῶν λευκῶν οὔτε γεύσει οὔτε ὄψει οὔτε ἀμφοῖν, ἀλλά τινι κοινῷ μορίῳ | τῶν αἰσθητηρίων ἁπάντων: ἔστι μὲν γὰρ μία αἴσθησις, καὶ τὸ κύριον αἰσθητήριον ἕν, τὸ δ᾽ εἶναι αἰσθήσει τοῦ γένους ἑκάστου ἕτερον, οἷον ψόφου καὶ χρώματος: τοῦτο δ᾽ ἅμα τῷ ἁπτικῷ μάλιστα ὑπάρχει: τοῦτο μὲν γὰρ χωρίζεται τῶν ἄλλων αἰσθητηρίων, τὰ & ἄλλα τούτου ἀχώριστα: εἴρηται δὲ | περὶ αὐτῶν ἐν τοῖς περὶ ψυχῆς θεωρήμασιν: φανερὸν τοίνυν ὅτι τούτου ἐστὶ πάθος ἡ ἐγρήγορσις καὶ ὁ ὕπνος. διὸ καὶ πᾶσιν ὑπάρχει τοῖς ζῴοις" καὶ γὰρ ἡ ἁφὴ μόνη πᾶσιν. εἰ γὰρ τῷ πάσας τι πεπονθέναι τὰς αἰσθήσεις καθεύδειν, ἄτοπον εἰ αἷς οὔτε ἀνάγκη οὔτε δυνατὸν | ἐνεργεῖν ἅμα, ταύτας ἀναγκαῖον ἀργεῖν ἅμα καὶ τοὐναντίον γὰρ εὐλογώτερον συνέβαινεν ἂν αὐταῖς, ἠρεμεῖν. ὡς δὲ νῦν λέγομεν, εὐλόγως ἔχει καὶ περὶ γὰρ κυρίου τῶν ἄλλων πάντων αἰσθητηρίου, καὶ πρὸς

τἄλλα,

πεπονθότος

πάντα,

ἐκείνων

τι δέ

συμπάσχειν τινος

ἀναγκαῖον

ἀδυνατοῦντος

οὐκ

ἐγίγνετο τὸ τρόπον τινὰ ἀκινητίζειντὸ μὴ ἅμα τούτων τοῦ ὃ συντείνει

καὶ

τὰ

λοιπὰ

ἀνάγκη

τοῦτ᾽

20

25

30

455b

ἀδυνατεῖν. φανερὸν δὲ ἐκ πολλῶν ὅτι οὐκ ἐν τῷ τὰς αἰσθήσεις ἀργεῖν καὶ μὴ χρῆσθαι αὐταῖς ὁ ὕπνος, οὐδ᾽ ἐν τῷ μὴ δύνασθαι αἰσθάνεσθαι: καὶ γὰρ ἐν tats | λειποψυχίαις τοιοῦτόν τι συμβαίνει: ἀδυναμία γὰρ αἰσθήσεως ἡ λειποψυχία, γίγνονται δὲ καὶ ἔκνοιαί τινες τοιαῦται: ἔτι δ᾽ οἱ τὰς ἐν τῷ αὐχένι φλέβας καταλαμβανόμενοι ἀναίσθητοι γίγνονται: ἀλλ᾽ ὅταν ἡ ἀδυναμία τῆς χρήσεως μήτ᾽ ἐν τῷ τυχόντι αἰσθητηρίῳ, μήτε δι᾽ ἣν ἔτυχεν αἰτίαν,

ἀλλά,

| καθάπερ

εἴρηται

νῦν,

ἐν

τῷ

πάντων: ὅταν μὲν γὰρ τοῦτ᾽ ἀδυνατήσῃ, αἰσθητηρίοις πᾶσιν ἀδυνατεῖν αἰσθέσθαι, ὅταν ἀνάγκῃ τούτῳ.

πρώτῳ



αἰσθάνεται

ἀνάγκη καὶ τοῖς δ᾽ ἐκείνων τι, οὐκ

10

ON SLEEP

455a

20

25

30

69

that one is seeing or hearing. For it is not sight, after all, whereby one sees that one is seeing. Nor is it taste or sight, or both, whereby one judges, or can judge, that sweet things differ from pale ones. One does so, rather, with some part that is common ! to all the sense-organs. For there exists a single sense-faculty, and the master sense-organ 1s single, though its being differs for the perception of each kind of thing, e.g. of sound or colour. And this accrues to an animal at the same time as the sense of touch, in particular. For this can exist separately from the other sense-organs, whereas the others are inseparable from it. These points have been discussed | in our studies on the soul. Given these things, it must plainly follow that waking and sleep are an affection of this. That is why they belong to all animals. For it is only touch that belongs to all of them. 455227. For if the sleeping state occurred because of some affection of all the senses, it would be odd that capacities that need not, and in a certain way cannot, | be exercised simultaneously, should necessarily be inactive and immobilized simultaneously. For just the opposite would be a more logical outcome, i.e. that they should not be simultaneously at rest. But on our present theory, a logical account can be given of them too. For when the sense-organ that controls all the others, and upon which the others

455b converge,

has undergone some affection, then | all the rest must be affected with it; whereas if any one of the latter is disabled, the former

need not be disabled as well. 7. 455b2-13. Sleep does not consist simply in loss of sensory awareness. For that occurs in several other states as well. Sleep is not the incapacitation of this or that special sense, but the disabling of the primary sense-organ.

10

455b2. It is plain, though, from many considerations that sleep does not consist merely in the senses being inactive and unused. Nor even in the incapacity to experience perception. For something similar happens in | fainting spells. For fainting is an incapacitation of perception. Certain unconscious states also take that form. Again, people who have pressure applied to the veins in the neck become insensible. Sleep occurs, rather, when the incapacity for use happens not in just any sense-organ, nor from just any cause, but, | as was just said, when it happens in the primary sense-organ whereby one perceives everything. For when that is disabled, then all of the sense-organs must be unable to perceive as well; whereas when one of the latter is disabled, the former need not be.

70

ΠΕΡῚ δι᾽

πάθος

ἣν

δ᾽

αἰτίαν

YIINOY συμβαίνει

KAI τὸ

ETPHTOPZEOZ καθεύδειν,

455b Kal

ποῖόν

τι

τὸ

γὰρ

ἐστί, λεκτέον. ἐπεὶ δὲ τρόποι πλείους τῆς αἰτίας | — καὶ τὸ τίνος ἕνεκεν, καὶ ὅθεν ἡ ἀρχὴ τῆς κινήσεως, καὶ τὴν

ὕλην

καὶ τὸν λόγον

αἴτιον

εἶναί φαμεν

-- πρῶτον

μὲν οὖν ἐπειδὴ

λέγομεν τὴν φύσιν ἕνεκά Tou ποιεῖν, τοῦτο δὲ ἀγαθόν τι, τὴν δ᾽ ἀνάπαυσιν παντὶ τῷ πεφυκότι κινεῖσθαι, μὴ δυναμένῳ δ᾽ ἀεὶ καὶ συνεχῶς κινεῖσθαι μεθ᾽ ἡδονῆς, ἀναγκαῖον εἶναι καὶ ὠφέλιμον, τῷ δὲ ὕπνῳ αὐτῇ Tfj ἀληθείᾳ προσάπτουσι τὴν μεταφορὰν ταύτην

ὡς

ἀναπαύσει

ὑπάρχει. ἡ 8° φρονεῖν πᾶσι ταῦτα, τὸ δὲ ζῴων ὑπάρχειν ὅτι εἰ ζῷον ὑπάρχειν αὐτῷ

ὄντι-

ὥστε

σωτηρίας

ἕνεκα

τῶν

7.

20

ζῴων

ἐγρήγορσις τέλος: τὸ γὰρ αἰσθάνεσθαι καὶ τὸ τέλος οἷς ὑπάρχει θάτερον αὐτῶν. βέλτιστα γὰρ τέλος | βέλτιστον, ὥστὲ ἀναγκαῖον ἑκάστῳ τῶν τὸν ὕπνον. λέγω δ᾽ ἐξ ὑποθέσεως τὴν ἀνάγκην, ἔσται ἔχον τὴν αὑτοῦ φύσιν, ἐξ ἀνάγκης τινὰ δεῖ, καὶ τούτων ὑπαρχόντων ἕτερα ὑπάρχειν.

ἔτι δὲ ποίας κινήσεως καὶ πράξεως ἐν τοῖς σώμασι γιγνομένης συμβαίνει τό [ τε ἐγρηγορέναι καὶ τὸ καθεύδειν τοῖς ζῴοις, μετὰ ταῦτα λεκτέον. τοῖς μὲν οὖν ἄλλοις ζῴοις καθάπερ τοῖς ἐναΐμοις ὑποληπτέον εἶναι τὰ αἴτια τοῦ πάθους ἢ ταὐτὰ ἢ τὰ ἀνάλογον, τοῖς δ᾽ ἐναίμοις ἅπερ τοῖς ἀνθρώποις: ὥστε ἐκ τούτων πάντα θεωρητέον.

6.

15

The text and exact sense at 455b20-21 are uncertain. The translation reads μεταφορὰν with Ross. Reading καταφορὰν with E, there is no reference to ‘metaphor’, but only to the 'downward movement that causes or constitutes sleep (456b24-32). The sense of αὐτῇ τῇ ἀληθείᾳ is unclear. Beare gives ‘taught by experience’; Hett, ‘in the light of the facts’; Barnes, ‘taught by truth itself. The present translation, ‘reflects the literal truth’, takes the point to be that the use of ἀνάπαυσις as a metaphor for sleep is well suited to the fact that sleep involves cessation from movement or activity. In a somewhat similar way, the English ‘rest’ signifies absence of movement or activity, and although not synonymous with ‘sleep’, is used in a special way of animals. They are capable of ‘resting’ in a way that inanimate objects, even when stationary, are not. The text gives no definite article at 455b23 before the first occurrence of τέλος, though several mss read τὸ τέλος. Whatever the correct text, translators have rightly understood the waking state, and the perceiving or thinking that occurs in it, not merely as an end but as the end. Thus Beare, ‘the highest end’; Hett, ‘the proper end'; Barnes, ‘the goal". Perceiving and thinking are the end for which the animal exists, since they are 'best' (024), constitutive of its being the kind of animal that it is. Its waking state is called the ‘end’, in that its self-realization through perceiving and thinking must occur while it is in that state.

25

30

4550

ON SLEEP

71

8. 455b13—34. What accounts for this? Four kinds of explanatory factor are recognized: (1) the ‘for the sake of something’; (2) the ‘source of change"; (3) the 'material'; and (4) the ‘account’. Sleep occurs for the sake of preserving the animal, by enabling it to exercise its capacities for perception and thought. If the animal is to exist as such, sleep is necessary for it to do so. 15

20

25

30

455b13. We must next give the explanation of sleeping, and say what kind of affection it is. Now there are several types of explanatory factor. | For we take as explanatory the ‘for the sake of something’, the ‘source of change’, the 'material', and the 'account'. First, then, given our claim that nature acts for the sake of something, and that that is a good, and that rest is both necessary and beneficial for everything that moves by nature, yet cannot move all the time and continuously with pleasure; | given also that metaphorical talk of sleep as 'rest' reflects the literal truth; it follows that sleep is for the sake of animal preservation. Waking, on the other hand, is the end.’ For perceiving and thinking are the end of all creatures to which either of them belongs. For they are best, and the end is | what is best. Hence? it is necessary for sleep to belong to every animal. I refer to 'conditional necessity', meaning that if there is to be an animal possessing its own nature, then certain things must belong to it of necessity; and if those are to belong, then so must certain others. 455b28. It must next be stated, moreover, from what kind of action occurring in their bodies | waking and sleeping arise must be assumed that the factors explaining the affection in are the same as, or analogous to, those for sanguineous ones; for sanguineous ones as for man. Hence all cases must be light of these.

8.

movement or in animals. It other animals and the same studied in the

The translation requires a comma or semicolon after βέλτιστον at 455b25, and ὥστε (the reading of some mss) rather than ἔτι δ᾽ (Ross). Thus also Beare and Hett. The effect of this is to make the whole passage b22-28 a continuous exposition of the final cause of sleep: b22-26 explains why 'it is necessary for sleep to belong to every animal’, and b26-28 explains what kind of necessity is relevant. With Ross's text, we must translate b25-26, 'Moreover, it is necessary for sleep to belong to every animal', and take the sentence as providing an additional argument for what has preceded it. Thus, Barnes gives 'Again, sleep belongs of necessity to each animal’. But the sentence does not provide any new argument for the thesis that sleep is for the sake of animal preservation.

72

ΠΕΡῚ

YIINOY

ΚΑΙ

455b

ΕΓΡΗΓΟΡΣΕΩΣ

ὅτι μὲν οὖν ἡ τῆς αἰσθήσεως ἀρχὴ γίγνεται [ ἀπὸ τοῦ αὐτοῦ μέρους τοῖς ζώοις ἀφ᾽ οὗπερ καὶ ἡ τῆς κινήσεως, διώρισται πρότερον ἐν ἑτέροις. αὕτη δέ ἐστι τριῶν διωρισμένων τόπων ὁ μέσος κεφαλῆς καὶ τῆς κάτω κοιλίας. τοῖς μὲν οὖν ἐναίμοις τοῦτ᾽ ἐστὶ τὸ περὶ τὴν καρδίαν μέρος. πάντα γὰρ | τὰ ἔναιμα καρδίαν ἔχει, καὶ ἡ ἀρχὴ τῆς κινήσεως καὶ τῆς αἰσθήσεως τῆς κυρίας ἐντεῦθέν ἐστιν. τῆς μὲν οὖν κινήσεως φανερὸν ὅτι καὶ ἡ τοῦ πνεύματος ἀρχὴ καὶ ὅλως ἡ τῆς καταψύξεώς ἐστιν ἐνταῦθα, καὶ τὸ ἀναπνεῖν τε καὶ τὸ τῷ ὑγρῷ καταψύχεσθαι πρός γε τὴν σωτηρίαν τοῦ ἐν τούτῳ μορίῳ θερμοῦ ἡ φύσις πεπόρικεν: ῥηθήσεται δὲ περὶ αὐτῆς ὕστερον καθ᾽ αὑτήν. τοῖς δὲ ἀναίμοις καὶ τοῖς ἐντόμοις και μὴ δεχομένοις πνεῦμα ἐν τῷ ἀνάλογον τὸ σύμφυτον πνεῦμα ἀναφυσώμενον καὶ συνιζάνον φαίνεται. δῆλον δὲ τοῦτο ἐπὶ τῶν ὁλοπτέρων, οἷον σφηκῶν καὶ μελιττῶν, καὶ ἐν ταῖς μυίαις | καὶ ὅσα τοιαῦτα. ἐπεὶ δὲ κινεῖν μέν τι ἢ ποιεῖν ἄνευ ἰσχύος ἀδύνατον, ἰσχὺν δὲ ποιεῖ ἡ τοῦ πνεύματος καθεξις, τοῖς μὲν εἰσφερομένοις ἡ θύραθεν, τοῖς δὲ μὴ ἀναπνέουσιν ἡ σύμφυτος (διὸ καὶ βομβοῦντα φαίνεται τὰ πτερωτά, ὅταν κινῆται, τῇ τρίψει τοῦ πνεύματος προσπίπτοντος πρὸς τὸ ὑπόζωμα τῶν ὁλοπτέρων), κινεῖται δὲ πᾶν αἰσθήσεώς τινος γενομένης, ἢ οἰκείας ἢ ἀλλοτρίας, ἐν τῷ πρώτῳ αἰσθητηρίῳ-εἰ δή ἐστιν ὁ ὕπνος καὶ ἡ ἐγρήγορσις πάθη τοῦ μορίου τούτου, ἐν ᾧ μὲν τόπῳ καὶ ἐν à μορίῳ πρώτῳ γίγνεται ὁ ὕπνος καὶ ἡ ἐγρήγορσις, φανερόν. 3

10

15

20

^

κινοῦνται δ᾽ ἔνιοι καθεύδοντες καὶ ποιοῦσι πολλὰ ἐγρηγορικά, οὐ μέντοι ἄνευ φαντάσματος καὶ αἰσθήσεώς τινος: τὸ γὰρ ἐνύπνιόν ἐστιν αἴσθημα τρόπον Tid: λεκτέον δὲ περὶ αὐτῶν ὕστερον. διότι δὲ τὰ μὲν ἐνύπνια μνημονεύουσιν ἐγερθέντες, τὰς δ᾽ ἐγρηγορικὰς πράξεις ἀμνημονοῦσιν, ἐν τοῖς Προβληματικοῖς εἴρηται.

9.

456a

At 456a20 some mss read γινομένης. This would make the animal's movement accompany, rather than follow, the perception. But the sentence makes most sense if the perception is thought of as preceding and causing the movement. See note 9 on 455b34-456a24 (on a20-21).

25

455b

ON SLEEP

73

9. 455b34—456a24. The 'source of change’ lies in the central organ of perception and movement. For sanguineous animals, this is the heart. Sleep and waking must therefore originate from there. 456a

15

20

455b34. Now it has already been determined in other works that perception in animals originates | from the same part as does movement. Of three areas that have been determined, this is the one that lies midway between the head and the lower abdomen. In sanguineous animals, this is the region about the heart. For all | sanguineous animals possess a heart, and both movement and perception in the full sense originate from there. As for movement, it is plain that breathing, and the cooling process in general, originate there; and that nature has provided both respiration and cooling by moisture with a view to conservation | of heat in that part. We shall discuss that subject later in its own right. But in bloodless animals and insects, and in those that do not respire, their naturally inherent breath can be seen inflating and contracting in the region analogous to the heart. This is clear in the case of whole-winged insects, such as wasps and bees as well as in flies | and all such. 456a15. Now, given that without strength it is impossible to move anything or to do anything, and the holding of breath produces strength (for things that inhale, holding of breath from outside; but for things that do not respire, holding of their inherent breath — which is why the winged insects of the whole-winged species can be heard buzzing when they move, because of the friction of breath impinging on their diaphragm); | and given that it is upon the occurrence of some perception, either internal or external, in the primary sense-organ, that every animal moves;? then, if sleep and waking are indeed affections of that part, the location and the bodily part in which sleep and waking first develop will be evident. 10. 456a24-29. Movements in sleep and dreams will be discussed later.

25

456a24. Some sleepers move and perform | many actions akin to waking ones, although not without some appearance (phantasma) or perception. For a dream is, in a certain way, a sense-impression. But those matters we must discuss later. Why it is that on being awakened people remember their dreams, but do not remember the actions akin to waking ones they have done, has been stated in the Problems.

74

ΠΕΡῚ

YIINOY

ΚΑΙ

ἘΓΡΗΓΟΡΣΕΩΣ

456a

8. Ἐχόμενον δὲ τῶν εἰρημένων ἐστὶν ἐπελθεῖν τίνων γιγνομένων καὶ πόθεν ἡ ἀρχὴ τοῦ πάθους γίγνεται, τοῦ T’ ἐγρηγορέναι καὶ τοῦ καθεύδειν. φανερὸν δὴ ὅτι ἐπεὶ ἀναγκαῖον τῷ ζῴῳ, ὅταν αἴσθησιν ἔχῃ, τότε πρῶτον τροφήν τε λαμβάνειν καὶ αὔξησιν, τροφὴ 8° ἐστὶ πᾶσιν ἡ ἐσχάτη τοῖς μὲν | ἐναίμοις ἡ τοῦ αἵματος φύσις, τοῖς δ᾽ ἀναίμοις τὸ ἀνάλογον, | τόπος δὲ τοῦ αἵματος αἱ φλέβες, τούτων δ᾽ ἀρχὴ ἡ καρδία (φανερὸν δὲ τὸ λεχθὲν ἐκ τῶν ἀνατομῶν) — τῆς μὲν οὖν θύραθεν τροφῆς εἰσιούσης εἰς τοὺς δεκτικοὺς τόπους γίγνεται ἡ ἀναθυμίασις εἰς τὰς φλέβας, ἐκεῖ δὲ μεταβάλλουσα ἐξαιμαίτοῦται καὶ πορεύεται ἐπὶ τὴν ἀρχήν. εἴρηται δὲ περὶ τούτων ἐν τοῖς περὶ τροφῆς νῦν δὲ ἀναληπτέον ὑπὲρ αὐτῶν τούτου χάριν, ὅπως τὰς ἀρχὰς τῆς κινήσεως θεωρήσωμεν, καὶ τί πάσχοντος τοῦ μορίου τοῦ αἰσθητικοῦ συμβαίνει ἡ ἐγρήγορσις καὶ ὁ ὕπνος. οὐ γάρ ἐστιν ὁ ὕπνος ἡτισοῦν ἀδυναμία τοῦ αἰσθητικοῦ, καθάπερ εἴρηται: καὶ yàp ἔκνοια καὶ πνιγμός τις καὶ λειποψυχία ποιεῖ τὴν τοιαύτην ἀδυναμίαν. ἤδη δὲ γεγένηταί τισι καὶ φαντασία λειποψυχήσασιν ἰσχυρῶς. τοῦτο μὲν οὖν ἔχει τινὰ ἀπορίαν: εἰ γὰρ ἐνδέχεται καταδαρθεῖν τὸν λειποψυχήσαντα, ἐνδέχοιτ᾽ ἂν ἐνύπνιον εἶναι καὶ τὸ | φάντασμα. πολλὰ δ᾽ ἐστὶν ἃ λέγουσιν οἱ σφόδρα λειποψυχήσαντες καὶ δόξαντες τεθνάναι: περὶ ὧν τὸν αὐτὸν λόγον ὑποληπτέον εἶναι πάντων. ἀλλὰ γάρ, ὥσπερ εἴπομεν, οὐκ ἔστιν ὁ ὕπνος ἀδυναμία πᾶσα τοῦ αἰσθητικοῦ, ἀλλ᾽ ἐκ τῆς περὶ τὴν τροφὴν ἀναθυμιάσεως γίγνεται τὸ πάθος τοῦτο: | ἀνάγκη γὰρ τὸ ἀναθυμιώμενον μέχρι 3

10.

^

4

[fd

M

*

LA



«

3

LA

said (455b4—6,

456b9-11),

35 456b

10

15

A

‘For if it is possible for one who has fainted to have fallen asleep’ (456b13-14).This translation takes the aorist infinitive καταδαρθεῖν, matching the aorist participle λειποψυχήσαντα, to refer to falling asleep as supposedly identical (or at least concomitant) with fainting, rather than as ensuing upon that state. Beare gives ‘if it were conceivable that one who has swooned should in this state fall asleep’ (italics added). Hett: ‘if it is possible for one who has fainted to fall asleep’. But there would be no reason for Aristotle to hypothesize that a subject who had fainted might subsequently fall asleep. What he needs to hypothesize is that someone who has fainted has already ipso facto fallen asleep. On that hypothesis it would be plausible to regard his experience, while in a faint, as a dream. But the hypothesis is incorrect, as has already twice been

30

and its rejection is

recalled once more at b17-18: ‘As we have said, however, sleep is not any and every incapacitation of the perceptual part’. With these words, the conditional is, in effect, pronounced counterfactual, and the dream solution thereby ruled out.

20

4568

ON SLEEP

75

Chapter Three 11. 456a30-b28. Sleep arises from the intake of food. Ingested matter undergoes exhalation, and rises in a dense mass, making the head heavy. Sleep ensues upon its return downward. 30

456a30. Following upon what has been said, it remains to consider, for both waking and sleeping, what developments give rise to the affection, and whence it has its origin. Now since it is when an animal possesses perception that it must first take in food and gain growth; and since for all | sanguineous animals food in its ultimate form is the natural substance

35 456b blood, and its counterpart for bloodless ones; | and since blood is located

10

15

20

in the veins, and these originate in the heart (a point that 1s clear from the dissections); it is clear that upon entry of food from an outside source into the parts fitted to receive it, the exhalation proceeds into the veins; and on undergoing change there, it is transformed into | blood, and travels to their point of origin. These matters have been discussed in our work on nutrition. We must now review them, for the purpose of studying the sources of change, and seeing what affection of the perceptual part gives rise to waking and sleep. For, as has been said, sleep is not just any incapacitation of the perceptual | part: unconsciousness, choking of a certain sort, and fainting produce such incapacitation also. And fantasy (phantasia) also has been known to occur in some people when they have fainted profoundly. 456b12. This point actually contains a certain difficulty. For if it is possible for one who has fainted to have fallen asleep,!? it could also be that the appearance (phantasma) was a | dream. Many things, too, are reported by those who have gone into a deep faint and have seemed dead. One must assume that the same account applies to all those cases. 456b17. Yet, as we have said, sleep is not any and every incapacitation of the perceptual part. Rather, this affection develops from the exhalation attendant upon nutrition. | For as the matter is exhaled, it has to be driven

76

ΠΕΡῚ

YIINOY

KAI

456b

ἘΓΡΗΓΟΡΣΕΩΣ

του ὠθεῖσθαι, εἶτ᾽ ἀντιστρέφειν καὶ μεταβάλλειν καθάπερ εὔριπον. τὸ δὲ θερμὸν ἑκάστου τῶν ζῴων πρὸς τὸ ἄνω πέφυκε φέρεσθαι: ὅταν δ᾽ ἐν τοῖς ἄνω τόποις γένηται, ἀθρόον πάλιν ἀντιστρέφει καὶ καταφέρεται. διὸ μάλιστα γίγνονται ὕπνοι ἀπὸ τῆς τροφῆς: ἀθρόον γὰρ πολὺ τό τε ὑγρὸν καὶ τὸ σωματῶδες ἀναφέρεται. ἱστάμενον μὲν οὖν βαρύνει καὶ ποιεῖ νυστάζεινὅταν δὲ ῥέψῃ κάτω καὶ ἀντιστρέψαν ἀπώσῃ τὸ θερμόν, τότε γίγνεται ὁ ὕπνος καὶ τὸ ζῷον καθεύδει. σημεῖον δὲ τούτων καὶ τὰ ὑπνωτικά: πάντα γὰρ καρηβαρίαν ποιεῖ, καὶ τὰ ποτὰ καὶ τὰ βρωτά, μήκων, μανδραγόρας, οἶνος, αἷραι. καὶ καταφερόμενοι kal νυστάζοντες τοῦτο δοκοῦσι πάσχειν, καὶ ἀδυνατοῦσιν αἴρειν τὴν κεφαλὴν καὶ τὰ βλέφαρα. καὶ μετὰ τὰ σιτία μάλιστα τοιοῦτος ὁ ὕπνος: πολλὴ γὰρ ἡ ἀπὸ τῶν σιτίων ἀναθυμίασις. ἔτι δ᾽ ἐκ κόπων ἐνίων: ὁ μὲν γὰρ κόπος συντηκτικόν, τὸ δὲ σύντηγμα γίγνεται ὥσπερ τροφὴ dinentos, àv μὴ ψυχρὸν ἧ. καὶ νόσοι δέ τινες τὸ αὐτὸ τοῦτο ποιοῦσιν, ὅσαι ἀπὸ περιττώματος ὑγροῦ καὶ θερμοῦ, οἷον συμβαίνει τοῖς πυρέττουσι καὶ ἐν τοῖς ληθάργοις. ἔτι & ἡ πρώτη ἡλικία: τὰ γὰρ παιδία καθεύδει σφόδρα διὰ τὸ τὴν | τροφὴν ἄνω φέρεσθαι πᾶσαν. σημεῖον δὲ τὸ ὑπερβάλλειν τὸ μέγεθος τῶν ἄνω πρὸς τὰ κάτω κατὰ τὴν πρώτην ἡλικίαν, διὰ τὸ ἐπὶ ταῦτα γίγνεσθαι τὴν αὔξησιν. διὰ ταύτην δὲ τὴν αἰτίαν καὶ ἐπιληπτικὰ γίγνεται: ὅμοιον γὰρ ὁ ὕπνος ἐπιλήψει, καὶ ἔστιν τρόπον τινὰ ὁ ὕπνος ἐπίληψις. διὸ καὶ συμβαίνει τισὶν ἡ ἀρχὴ τούτου τοῦ πάθους καθεύδουσιν, καὶ καθεύδοντες μὲν ἁλίσκονται, ἐγρηγορότες δ᾽ οὔ - ὅταν γὰρ πολὺ φέρηται τὸ πνεῦμα

τὸν

ἄνω,

πόρον

συμφέρουσιν

καταβαῖνον

8U

οἱ

oU



οἶνοι,

πάλιν

τὰς

ἀναπνοὴ

οὐδὲ

ταῖς

φλέβας

γίγνεται.

| τίτθαις

ὀγκοῖ,

διὸ

καὶ

25

30

35 457a

10

συνθλίβει

τοῖς

παιδίοις

(διαφέρει

γὰρ

οὐ

ἴσως

οὐδὲν αὐτὰ πίνειν ἢ τὰς τίτθας), ἀλλὰ δεῖ πίνειν ὑδαρῆ καὶ ὀλίγον: πνευματῶδες γὰρ ὁ οἶνος καὶ τούτου μᾶλλον ὁ μέλας.

15

ON SLEEP

456b

25

71

to a certain point, and must then turn back and change direction, like the tide in a narrow channel. Now in every animal the hot matter tends naturally to move upwards. But when it has reached the upper areas, it turns back again, and comes down in a mass. That is why spells of sleep follow especially upon the intake of food: | it is because the moist and solid matter are rising in a dense mass. This, while static, weighs one down and causes nodding. But when it has descended again, and by returning has repelled the hot matter, then sleep ensues and the animal falls asleep. 12. 456b28-457a3. Soporific agents have the same effect, as do fatigue and certain illnesses.

456b28. Soporific agents are a further proof of this. For all of them, liquid land solid alike (poppy, mandragora, wine, darnel), produce heaviness of the head. And people seem to be affected in that way when sinking into sleep or nodding off: they are unable to hold up their heads or eyelids. It is after meals, especially, that sleep of that sort comes on. For the exhalation from food is considerable. It also follows upon some states of fatigue. For 35 fatigue | is a solvent; and the dissolved matter, unless it is cold, acts like 457a | food not yet digested. Certain illnesses too have the same effect, i.e. those due to moist and hot residue, as happens with the feverish and the 30

comatose. 13. 457a3-21. The same process underlies the sleep of infants. It also explains epileptic seizures, and the harmful effects of wine upon infants. 5

10

15

45723. Again, early childhood is similar. For infants sleep deeply because alltheir | food is traveling upwards. Proof of this is the fact that during early childhood the upper parts exceed the lower in size, because growth takes place in the direction of the former. This is also the explanation of epileptic fits. For sleep is like epilepsy. In fact, sleep is, in a way, an epileptic seizure. That is why, | for some people, the start of that affection happens when they are asleep; and they become seized with it while asleep, but not while awake. For whenever vapour moves upwards in a large volume, it swells the veins as it comes down again, and constricts the respiratory passage. Hence wines are not good for infants — or for their | wet-nurses, since it probably makes no difference whether they drink it themselves or their nurses do. But they should drink it diluted and in small quantity. For wine is gaseous, particularly dark wine. Infants are so filled

78

ΠΕΡῚ ts

N

οὕτω

δὲ

οὐδὲ €

x

τὰ

αἴτιον

ΥΠΝΟΥ

7

ἄνω

τὸν

H

αὐχένα:

2

καὶ

τοῦ

τὸ

ὅλον

δὲ

ἐν

γὰρ

Y

τὰ

γὰρ

αἱ

πέντε

καὶ

*

,

τὸ

τὸ

οἱ

^

μεθύουσιν,

ων

elvai

ἔμβρυα

φλέβες

μηνῶν

LA

σφόδρα

τοῦτ᾽

οἱ ἀδηλόφλεβοι

μὲν

dore A^.

δὲ

μήτραις

457a

iva

τοῖς x

εὔλογον

ταῖς

τῶν

παιδίοις, La

ὥσπερ ,

φίλυπνοι

μεγαλοκέφαλοι:

τοῖς

πολλή.

ἠρεμεῖν

ΕΓΡΗΓΟΡΣΕΩΣ ^

τροφῆς

| ἀναφέρεται

καὶ

ΚΑΙ

^

πλήρη

στρέφουσι ΄

ὑγρότης

οἱ

LA

πάθος

20

πρῶτον.

νανώδεις

στεναί,

καὶ

ὥστ᾽

οὐ

ῥάδιον διαρρεῖν κατιὸν τὸ ὕγρόν, τοῖς δὲ νανώδεσι καὶ μεγαλοϊκεφάλοις ἡ ἄνω ὁρμὴ πολλὴ καὶ ἀναθυμίασις. οἱ δὲ φλεβώδεις οὐχ ὑπνωτικοὶ δι᾽ εὔροιαν τῶν πόρων, ἂν μή τι ἄλλο πάθος ἔχωσιν ὑπεναντίον. οὐδ᾽ οἱ μελαγχολικοί: κατέψυκται γὰρ ὁ

εἴσω

τόπος,

ὥστ᾽

οὐ

γίγνεται

πλῆθος

αὐτοῖς

25

ἀναθυμιάσεως.

διὰ τοῦτο γὰρ καὶ βρωτικοί, σκληφροὶ ὄντες: | ὥσπερ γὰρ οὐδὲν ἀπολελαυκότα διάκειται τὰ σώματα αὐτοῖς. ἡ δὲ μέλαινα χολὴ φύσει ψυχρὰ οὖσα καὶ τὸν θρεπτικὸν τόπον ψυχρὸν ποιεῖ καὶ τὰ ἄλλα μόρια, ὅπου ἂν ὑπάρχῃ δυνάμει τὸ τοιοῦτον περίττωμα. ὥστε

φανερὸν

ἐκ

τοῦ

θερμοῦ

εἴσω

t^

τις 2

*

Z

3

εἰρημένην »

,

x

7

διὸ

πολλῃ

^

Y

καὶ τὰ μέν κάτω θερμά,

3

ὅτι

Z

καὶ ^



+



κίνησις

x

καὶ

^

διὰ

,

Ϊ ἄνω κατέψυκται οἷον τὰ περὶ τοὺς

ἄλλα

ὕπνον

,

θερμότητα

εἶναι

v

Exovra

κατάψυξιν,

τ

ὕπνος

b

^

τοιαῦτα,

δ᾽

,

ig

Y

x

ἔστι

αἴτια

δ᾽

οὐκ

τοῦ

t^



x

457b 3

2

βλέφαρα,

Y

b

έντος

D

+

τὴν

ὅθεν

τὰ ;»

7

διὰ

^

ἄν, ὅτι μετὰ ἔστιν ὑπνωτικὰ

Y

,

ὑπνοῦντος.

καταπίπτει

x

σύνοδός

^

φυσικὴ

καὶ τὰ ἔξω, τὰ δ᾽ πόδας καὶ τὰ εἴσω.

^

τὰ

τοῦ

ἐστὶ

t

,

ψῦξιν

καίτοι τοῦτό τις ἀπορήσειεν ἰσχυρότατος ὁ ὕπνος γίγνεται, καὶ v



os

ἀντιπερίστασις

^

καταψύχεται x

εἰρημένων

^

bi

αἰτίαν.

ἐκλείπει, ^

[ τῶν

,

30

καὶ

^

τὰ

τὰ σιτία οἶνος καὶ

>

εὔλογον

καθεύδειν

>

x

τὸν

x

μὲν

θερμά.

20

79

ON SLEEP

4578

with food in their upper parts that for five months they do not even turn their necks. For as with the very drunk, much moisture | gets carried upwards. This affection probably also explains why initially the embryo lies at rest in the womb. 14. 457a21—33. The same pattern occurs in those with small veins or big heads. Those with large veins and atrabilious people show the opposite pattern. 457a21.

In general,

also, those

with

inconspicuous

veins,

as well

as

dwarfs and those with big heads, are fond of sleeping. For in the former 25

the veins are narrow, so that the descending moisture cannot easily flow through them; while for dwarfs and | big-headed creatures the surge upwards in exhalation is considerable. Those with large veins, on the other hand, because of the breadth of their passages, are not given to sleep, unless they possess some other, counteracting affection. Nor again are the atrabilious. For their inner region has been cooled, so that in their case no mass exhalation occurs. For the same reason they are much inclined to eat,

30

gaunt as they are. | For their bodily state is as if they have gained no benefit from their food. Black bile, also, being cold by nature, makes the nutrient area cool, and other parts too, wherever a secretion of that kind

may potentially be present. 15. 457a33-b6. Conclusion. Sleep is due to an inner concentration of hot matter and its subsequent reverse flow downward. 457b 457233. Hence it is plain from | what has been said that sleep is a sort of inward concentration of the hot matter and a natural reflux that are due to the causes stated. That is why the movement of a somnolent person is considerable. But from where the heat fails, he is cooled; and owing to the 5 cooling his eyelids droop. Both upper and outer parts | have been cooled, while inner and lower ones, e.g. those around the feet and the interior, are hot. 16. 457b6—26. A problem. If the things that produce sleep are hot, why should sleep itself be a cooling process? Some possible solutions.

10

457b6. One might, however, pose the following problem. It is after food that sleep is heaviest; and wine and other such things containing heat are soporific. Yet it is not logical that sleep should be a | cooling process,

80 πότερον

ΠΕΡῚ οὖν

τοῦτο

YIINOY

συμβαίνει

ΚΑΙ ὅτι

4570

ἘΓΡΗΓΟΡΣΕΩΣ ὥσπερ



κοιλία

κενὴ

μὲν

οὖσα

θερμή ἐστιν, ἡ δὲ πλήρωσις αὐτὴν καταψύχει διὰ τὴν κίνησιν, οὕτω καὶ οἱ ἐν τῇ κεφαλῇ πόροι καὶ τόποι καταψύχονται ἀναφερομένης τῆς ἀναθυμιάσεως; ἢ ὥσπερ τοῖς προσχεομένοις τὸ | θερμὸν ἐξαίφνης φρίκη γίγνεται, κἀκεῖ ἀνιόντος τοῦ θερμοῦ ἀθροιζόμενον τὸ ψυχρὸν καταψύχει, καὶ τὸ κατὰ φύσιν θερμὸν ποιεῖ ἐξαδυνατεῖν καὶ ὑποχωρεῖν; ἔτι δὲ πολλῆς ἐμπιπτούσης τροφῆς, ἣν ἀνάγει τὸ θερμόν, ὥσπερ τὸ πῦρ ἐπιτιθεμένων τῶν ξύλων, καταψύχεται, ἕως ἂν καταπεφθῇ. | γίγνεται γὰρ ὁ ὕπνος, ὥσπερ εἴρηται, τοῦ σωματώδους ἀναφερομένου ὑπὸ τοῦ θερμοῦ διὰ τῶν φλεβῶν πρὸς τὴν κεφαλήν: ὅταν δὲ μηκέτι δύνηται, ἀλλὰ τῷ πλήθει ὑπερβάλλῃ τὸ ἀναχθέν, πάλιν ἀνταπωθεῖται καὶ κάτω ῥεῖ (διὸ καὶ πίπτουσί γε ὑποσπωμένου τοῦ θερμοῦ τοῦ ἀνάγοντος οἱ ἄνθρωϊποι: μόνον γὰρ ὀρθὸν τῶν ζῴων), καὶ ἐπιπεσὸν μὲν ἔκνοιαν ποιεῖ, ὕστερον δὲ φαντασίαν. ἢ αἱ μὲν νῦν λεγόμεναι λύσεις ἐνδεχόμεναι μέν εἰσι τοῦ γίγνεσθαι τὴν κατάψυξιν, οὐ μὴν ἀλλὰ κύριός γ᾽ ἐστὶν ὁ τόπος ὁ περὶ τὸν ἐγκέφαλον, ὥσπερ ἐν ἄλλοις εἴρηται. πάντων δ᾽ ἐστὶ τῶν ἐν τῷ σώματι | ψυχρότατον ὁ ἐγκέφαλος, τοῖς δὲ μὴ ἔχουσι τὸ ἀνάλογον τούτῳ μόριον. ὥσπερ οὖν τὸ ἀπατμίζον ὑγρὸν ὑπὸ τῆς τοῦ ἡλίου θερμότητος, ὅταν ἔλθῃ εἰς τὸν ἄνω τόπον, διὰ τὴν ψυχρότητα αὐτοῦ καταψύχεται καὶ συστὰν καταφέρεται | γενόμενον πάλιν ὕδωρ, οὕτως ἐν τῇ ἀναφορᾷ τοῦ θερμοῦ τῆ πρὸς τὸν ἐγκέφαλον ἡ μὲν περιττωματικὴ ἀναθυμίασις εἰς φλέγμα συνίσταται (διὸ καὶ οἱ κατάρροι φαίνονται γιγνόμενοι ἐκ τῆς κεφαλῆς), ἡ δὲ τρόφιμος καὶ μὴ νοσώδης | καταφέρεται συνισταμένη καὶ καταψύχει τὸ θερμόν. πρὸς δὲ τὸ καταψύχεσθαι καὶ μὴ δέχεσθαι ῥᾳδίως τὴν ἀναθυμίασιν συμβάλλεται καὶ ἡ λεπτότης καὶ στενότης τῶν περὶ τὸν ἐγκέφαλον φλεβῶν. τῆς μὲν οὖν καταψύξεως τοῦτ᾽ ἐστὶν αἴτιον, καίπερ τῆς ἀναθυμιάσεως ὑπερβαλλούσης τῇ | θερμότητι.

15

20

25

30

4588

10

4570

15

ON SLEEP

81

when the things that cause sleeping are hot. Is the explanation that, just as the stomach is hot when empty, and yet its filling cools it because of the movement, in the same way the passages and regions of the head are cooled as the exhalation moves upwards? Or is it that, just as | a sudden shiver comes over people who have hot water poured on them, so in this case, as the heat rises, the cool rallies and counteracts it, thereby causing

what is naturally hot to lose its power and to withdraw? Again, when much

20

food is ingested, and the heat carries it up, then the hot is cooled,

like a fire when logs are placed upon it, until digestion is complete. | For sleep comes on, as has been said, when the solid matter is carried up by

heat, through the veins, to the head. When the matter taken up can rise no

25

longer, but is excessive in bulk, it gets driven back again and flows downward. (That explains why human beings actually fall, when the uplifting heat is withdrawn. | For man alone among animals is erect). And as the solid matter descends, it produces unconsciousness, and subsequently fantasy (phantasia). 17. 457b26-458a10. An alternative solution.

The brain acts as a coolant

during sleep. 457b26. Alternatively, the solutions just stated are admissible explanations of the cooling; it is, nevertheless, the region about the brain that is the dominant factor, as has been said elsewhere. The brain, or its

30

counterpart in animals that do not have one, | is the coldest of all parts of the body. So, just as the moisture that is vaporized by the sun's heat when it reaches the upper region, is cooled by its coldness, and after condensing

458a comes down once more | in the form of water, in the same way, as the hot

matter rises to the brain, the superfluous exhalation gathers into phlegm (which is why catarrhs are felt as developing from the head), while the 5

nutritive

10

cools the hot. A further contribution to the cooling, and to its resistance to exhalation, comes from the fineness and narrowness of the veins in the head. This accounts for the cooling, then, despite the excess of heat in the | exhalation itself.

and

wholesome

matter

| collects

together,

travels down,

and

82

ΠΕΡῚ

ἐγείρεται θερμότης

ἐν

δ᾽

YIINOY

ὅταν

ὀλίγῳ

ΚΑΙ

πεφθῆῇῆ

πολλὴ

ἐκ

τοῦ

ἘΓΡΗΓΟΡΣΕΩΣ

καὶ

κρατήσῃ

περιεστῶτος,

458a

ἡ καὶ

συνεωσμένη διακριθῇ

τό

τε σωματωδέστερον αἷμα καὶ τὸ καθαρώτερον. ἔστι δὲ λεπτότατον μὲν αἷμα καὶ καθαρώτατον τὸ ἐν τῇ κεφαλῆ, παχύτατον δὲ καὶ θολερώτατον τὸ ἐν τοῖς κάτω | μέρεσιν. παντὸς δὲ τοῦ αἵματος ἀρχή, ὥσπερ εἴρηται καὶ ἐνταῦθα καὶ ἐν ἄλλοις, ἡ καρδία. τῶν δ᾽ ἐν τῇ καρδίᾳ ἑκατέρας τῆς θαλάμης κοινὴ ἡ μέση ἐκείνων δ᾽ ἑκατέρα δέχεται ἐξ ἑκατέρας τῆς φλεβός, τῆς τε μεγάλης καλουμένης καὶ τῆς ἀορτῆς. ἐν δὲ τῇ μέσῃ γίγνεται ἡ διάκρισις. | ἀλλὰ τὸ μὲν διορίζειν περὶ τούτων ἑτέρων ἐστὶ λόγων οἰκειότερον: διὰ δὲ τὸ γίγνεσθαι ἀδιακριτώτερον τὸ αἷμα μετὰ τὴν τῆς τροφῆς προσφορὰν ὕπνος γίγνεται, ἕως ἂν διακριθῇ τοῦ αἵματος τὸ μὲν καθαρώτερον εἰς τὰ ἄνω, τὸ δὲ θολερώτερον εἰς τὰ κάτω: ὅταν δὲ τοῦτο συμβῇ, ἐγείρονται ἀπολυθέντα τοῦ ἐκ τῆς τροφῆς βάρους. τί μὲν οὖν τὸ αἴτιον τοῦ καθεύδειν εἴρηται, ὅτι ἡ τοῦ σωματώδους τοῦ ἀναφερομένου ὑπὸ τοῦ συμφύτου θερμοῦ ἀντιπερίστασις ἀθρόως ἐπὶ τὸ πρῶτον αἰσθητήριον: καὶ τί ἐστιν ὁ ὕπνος, ὅτι τοῦ πρώτου αἰσθητηρίου κατάληψις πρὸς τὸ μὴ δύνασθαι ἐνεργεῖν, ἐξ | ἀνάγκης μὲν γινόμενος - οὐ γὰρ ἐνδέχεται



ἕνεκα

ἔῷον

δὲ

εἶναι

ooTnpíag-

μὴ

συμβαινόντων

σῴζει

γὰρ

τῶν

ἀπεργαζομένων

ἡ ἀνάπαυσις.

αὐτό

15

20

25

30

ON SLEEP

4588

83

18. 458a10-25. Awakening occurs when the more solid and the purer blood, which were intermingled during sleep, have become separated

once more. 458a10. Awakening occurs when digestion is complete, i.e. when the heat, which has been concentrated in large quantity from the surrounding area in a small space, prevails, and when separation has taken place between the more solid and the purer blood. The rarest and purest blood is located in the head, while the thickest and most turbid is in the lower 15

20

25

| parts. The source of all blood, as has been said, both here and elsewhere,

is the heart. Of the chambers in the heart, the central one is connected with both of the others; and each of these receives blood from one of the two blood-vessels, i.e. the so-called 'great vein' and the aorta. It is in the central chamber that the separation takes place. | Detailed treatment of these matters is, however, more proper to another treatise. But it is because the blood becomes more intermingled after the absorption of food that sleep occurs. It lasts until the purer kind of blood has been separated into the upper regions and the more turbid into the lower. Once that has happened, animals, ! released from the heaviness due to the intake of food, wake up. 19. 458a25-32. Retrospect. Sleep has been explained in terms of the material changes that produce it. It consists in a disabling of the primary sense-organ, and is necessary for animal preservation.

30

458225. The cause of sleeping has, then, been stated. It is the reverse flow of the solid matter, carried upwards by the naturally inherent heat, en masse towards the primary sense-organ. It has also been stated what sleep is; i.e. that it is a seizure of the primary sense-organ, rendering it incapable of functioning. It occurs of | necessity, since it is not possible for an animal to exist, should the conditions that produce it not obtain. Yet it is for the sake of animal preservation. For rest does preserve it.

84

458a

ΠΕΡῚ

ENYIINION

1. Μετὰ δὲ ταῦτα περὶ ἐνυπνίου ἐπιζητητέον, καὶ πρῶτον τίνι τῶν τῆς ψυχῆς φαίνεται, καὶ πότερον τοῦ νοητικοῦ τὸ πάθος ἐστὶ τοῦτο ἢ τοῦ αἰσθητικοῦ: τούτοις γὰρ μόνοις τῶν ἐν ἡμῖν γνωρίζομέν τι. 3

3

εἰ

t

^

Y

Uu

x

3

^

M

2

΄

458b

x

δ᾽ ἡ χρῆσις ὄψεως ὅρασις, Kal ἀκοῆς τὸ ἀκούειν, Kai ὅλως αἰσθήσεως τὸ αἰσθάνεσθαι, κοινὰ δ᾽ ἐστὶ τῶν αἰσθήσεων oiov σχῆμα καὶ μέγεθος καὶ κίνησις καὶ τἄλλα τὰ τοιαῦτα, ἴδια δ᾽ oiov χρῶμα ψόφος xupós, ἀδυνατεῖ δὲ πάντα μύοντα καὶ th

3

T

Fd

N

^

3

,

^

,

καθεύδοντα

vw

3

ἄνευ

3

opav,

ὁμοίως

^

t

μὴν ^

οὐδὲ

τοῖς

τῇ

+

^

,

\

λοιπῶν,

^

ἄρα

γὰρ b

,

M

φήσειεν,

ouT'

3

προσιόν 7

καλόν: ^

ὧν

»

ἀληθῶς

οὐκ

t

kJ

αἰσθήσει

la



3

τῇ

τὸ US

λευκὸν

ὅτι

^

γε

μόνον x

καὶ

δῆλον

^

οὐκ

οὐ

3

av

τῶν

N

ὕπνοις:

ἀλλὰ

Ww

ἐπὶ

^

δόξῃ:

εἶναι,

ou8év

Kal

x

T

ἵππον

FG

δὲ

΄

t



αἰσθήσεως

τὸ

φαμεν t

δόξα ^

οὔτε

έτι

παρὰ

ἐγρηγορέναι καὶ

TO

\

^

#

|

>

+

ἐνίοτε

ἐννοοῦμεν

#

αἰσθανόμενοί

διανοούμεθάα

φαντάσματα

ἐνύπνιον 2

TU vw

ἄλλα

2

^

τι.

περὶ

οὕτω

καὶ

|

+

Z^

οὗ

ἐννοοῦμεν.

a

ἐν

ἄλλο

τι,

v

γὰρ

ψευδῶς.

καθάπερ

^

φανείη

ὕπνοις δ᾽

᾿

bd

ἂν

TW

bi

ἐν

γὰρ 15

EV

if

αἰσθανόμεθα,

τοῖς

10





8€ τοῖς ὕπνοις συμβαίνει τὴν ψυχὴν τοῦτο ποιεῖν: ὁμοίως ὅτι ἄνθρωπος καὶ ὅτι λευκὸς ὁ προσιὼν δοκοῦμεν ὁρᾶν. Y

5

,

^



αἰσθανόμεθα οὐδὲν ἐν ἐνύπνιον αἰσθανόμεθα. ἀλλὰ

^

A^

^

5

Y

x

,

T

ἄνθρωπον

,

2

TH

~

πολλάκις παρὰ Z^

τοῦτο,

τὰ hia

EL

4588

85 ON DREAMS Chapter One 20. 458a33-b3. thought?

Introduction.

Is dreaming

the work of perception

or of

458b 458a33. Next we must inquire into the dream: first | with what part of the soul it is apparent, i.e. whether this affection is the work of the thinking or the

perceptual part. For of the faculties within us, it is with these alone that we know anything. 21. 458b3-9. It cannot be with perception that we are aware of dreams.

5

458b3. Given, then, that the exercise of sight is seeing, and that of hearing is hearing, and in general that of perception is perceiving — and the senses have | common objects, e.g. shape, size, movement and the like, and again special ones, e.g. colour, sound and taste — and given that everything with its eyes closed in sleep is unable to see, and likewise for the other senses as well, so that clearly we are not perceiving anything during periods of sleep; it follows that it 15 not with perception, anyway, that we perceive a dream. 22. 458b10—15. Nor are we aware of dreams with judgment. For in sleep an approaching figure is said to be both a man and pale, a pronouncement which judgment without perception could not make.

10

458b10. But neither is it with judgment that we do so. For we say not only that the thing approaching is a man or a horse, but also that it is pale or beautiful. None of those things could be said by judgment, either truly or falsely, without perception. Yet during periods of sleep the soul, in effect, does just that. For we judge that we see, equally, that the figure approaching is a man and that he is pale. 23. 458b15-25. Moreover, a dream is to be distinguished from thoughts in sleep. It is the latter that we cognize with judgment.

15

other

458b15. Again, over and above the dream we think something else, just as we do when perceiving something during the waking state. For when we are perceiving something, we often also engage in some thought about it. So too during periods of sleep we sometimes think other things over and above the appearances (phantasmata). 'This would be apparent to anyone who should

86

ΠΕΡῚ

ENYIINION

458b

τις προσέχοι τὸν νοῦν καὶ πειρῷτο μνημονεύειν [ ἀναστάς. ἤδη δέ τινες καὶ ἑωράκασιν ἐνύπνια τοιαῦτα, οἷον οἱ δοκοῦντες κατὰ τὸ μνημονικὸν παράγγελμα τίθεσθαι τὰ προβαλλόμενα' συμβαίνει γὰρ αὐτοῖς πολλάκις ἄλλο τι παρὰ τὸ ἐνύπνιον τίθεσθαι πρὸ ὀμμάτων εἰς τὸν τόπον φάντασμα: ὥστε δῆλον ὅτι οὐκ ἐνύπνιον πᾶν τὸ ἐν ὕπνῳ φάντασμα, καὶ ὅτι ὃ ἐννοοῦμεν τῇ δόξῃ

20

25

δοξάζομεν.

δῆλον δὲ περὶ τούτων ἁπάντων τό γε τοσοῦτον, ὅτι τὸ αὐτὸ ᾧ καὶ ἐγρηγορότες ἐν ταῖς νόσοις ἀπατώμεθα, τοῦτ᾽ αὐτὸ καὶ ἐν τῷ ὕπνῳ ποιεῖ τὸ πάθος. καὶ ὑγιαίνουσι δὲ καὶ εἰδόσιν ὅμως ὁ ἥλιος ποδιαῖος εἶναι δοκεῖ. t

ἀλλ᾽ εἴτε δὴ ταὐτὸν εἴθ᾽ | ἕτερον καὶ τὸ αἰσθητικόν, οὐδὲν ἧττον οὐ αἰσθάνεσθαί τι: τὸ γὰρ παρορᾶν καὶ τι καὶ ἀκούοντος, οὐ μέντοι τοῦτο ὑπόκειται μηδὲν ὁρᾶν | μηδ᾽ ἀκούειν ^

ap’ τὴν

τι

s.

5

ouv

αἴσθησιν

καὶ

τὰς

ἐγρηγορότος >



x

TO

x

pév οὐκ

μὴ



προσβάλλει 7

despite

x

μηδὲν

ἀλλ᾽

μέν A

καὶ ὁτὲ ὁτὲ δὲ

2

,

ἀληθές,

πως



^

x

τὸ

τὴν

δὲ

αἰσθήσει, 3

A

δὲ

καὶ

ἕκαστον

/,

4

μηδὲν

πάσχειν

ὄψιν

πάσχειν

τούτων οὐχ I

ὥσπερ οὕτω ita

δὲ

b

μὲν ἡ δόξα λέγει ὅτι ψεῦδος, κατέχεται καὶ ἀκολουθεῖ τῶ

follows Barnes

their separation,

x

ἐνδέχεται

αἰσθήσεις,

The translation at 458b22-24 φάντασμα,

^

ὁρᾶν

GANGES,

ἄλλας

ὥσπερ ἐγρηγορότος: ὥσπερ ἐγρηγορόσιν, φαντάσματι;

11.

M

τὸ φανταστικὸν τῆς ψυχῆς γίνεται ἄνευ τοῦ ὁρᾶν καὶ παρακούειν ὁρῶντος ἀληθές ὃ οἴεται. ἐν δὲ τῷ ὕπνῳ μηδ᾽ ὅλως αἰσθάνεσθαι.

as object

and Hett in taking ἄλλο of τίθεσθαι.

Beare

τι with

(note ad loc.)

evidently takes ἄλλο τι as subject of συμβαίνει, assumes a comma after ἐνύπνιον (023), and understands τίθεσθαι ... φάντασμα in apposition to ἄλλο Tu He translates: "They frequently find themselves engaged in something else besides the 12.

dream, viz. in setting a phantasm which they envisage into its mnemonic position'. ‘Declare it an illusion’ (45936). This translates Ross's text, but leaves it unclear, as at 460b15, what exactly is declared illusory. There is ms support for reading τὸ ὁρώμενον after ψεῦδος. Thus, Hett: ‘opinion states ... that what is seen is false’, Beare: 'the object seen is an illusion τὸ ὁρώμενον may be a gloss, but if so, it probably represents Aristotle's meaning, whether or not it was part of his original text.

30

459a

4580

20

25

ON DREAMS

87

make a mental effort to remember | on arising. In fact, some people have actually experienced such dreams, e.g. those who judge that they are arranging a given set of items according to the system for memorizing them. For they often find themselves putting into its place before their eyes some other appearance (phantasma) apart from the dream.!! Thus, it is clear that not every appearance (phantasma) during sleep is a dream, | and that it is what we are thinking that we cognize with judgment. 24. 458b25-29. If a dream is the work of neither perception nor judgment, what is it? This much is clear: a dream is due to the faculty by which we are perceptually deceived when awake but ill, or even when in good health. 458b25. On all these issues this much at least is clear: it is the very same faculty by which we are deceived during illnesses, even when awake, that also produces the affection during sleep. Why, even to those who are in good health and who know otherwise, the sun still seems to be only one foot

across. 25. 458b29-459a1. Yet mis-seeing, mis-hearing etc. require that we perceive something; whereas in sleep, it is being assumed, we are not perceiving anything. 458b29. But whether in fact the imagining (phantastikon) part of the soul and 30 the perceptual are the same | or different, still the affection does not occur without our seeing or perceiving something For someone who mis-sees or mis-hears must be seeing or hearing some real thing, albeit not the thing he 459a supposes. But during sleep, it is being assumed, one is neither seeing | nor hearing, nor, in general, perceiving anything. 26. 459a1-8. Possible solution: even though we are not perceiving anything, the senses may still be affected in some way. And judgment sometimes recognizes this experience as deceptive, but at other times accepts it.

5

459a1. Then could it be true that one is not seeing anything, yet not true that the sense is in no way affected? Is it possible, rather, that both sight and the other senses are affected somehow, | and that each of these impinges in some way upon perception, as with a waking person, though not in the same way as with a waking person? And does judgment sometimes declare it an illusion,!2 as it does for waking people, while at other times it is held in check and follows along with the appearance (phantasma)?

88 ὅτι

τὸ

πάθος

μὲν

οὖν

τοῦτο

| αἰσθανομένου

οὐκ

ΠΕΡῚ

ἘΝΥΠΝΊΩΝ

τοῦ

δοξάζοντος

ἔστι

ὃ καλοῦμεν

ἁπλῶς:

ὁρᾶν

ἐνυπνιάζειν,

γὰρ

4598 οὐδὲ

τοῦ

φανερόν.

ἂν ἦν καὶ ἀκούειν

διανοουμένου

ἀλλ᾽

οὐδὲ

ἁπλῶς.

τοῦ

ἀλλὰ

10

πῶς δὴ καὶ τίνα τρόπον, ἐπισκεπτέον. ὑποκείσθω μὲν οὖν, ὅπερ ἐστὶ καὶ φανερόν, ὅτι τοῦ αἰσθητικοῦ τὸ πάθος, εἴπερ καὶ ὁ ὕπνος: οὐ γὰρ ἄλλῳ μέν τινι τῶν ζῴων ὑπάρχει ὁ ὕπνος, ἄλλῳ δὲ τὸ ἐνυπνιάζειν, ἀλλὰ τῷ αὐτῷ. ἐπεὶ δὲ | περὶ φαντασίας ἐν τοῖς περὶ ψυχῆς εἴρηται, καὶ ἔστι μὲν τὸ αὐτὸ τῷ αἰσθητικῷ τὸ φανταστικόν, τὸ δ᾽ εἶναι φανταστικῷ καὶ αἰσθητικῷ ἕτερον, ἔστι δὲ φαντασία ἡ ὑπὸ τῆς κατ᾽ ἐνέργειαν αἰσθήσεως γινομένη κίνησις, τὸ 8° ἐνύπνιον φάντασμά τι φαίνεται εἶναι (τὸ γὰρ ἐν ὕπνῳ φάντασμα ἐνύπνιον λέγομεν, εἴθ᾽ ἁπλῶς εἴτε τρόπον τινὰ γινόμενον), φανερὸν ὅτι τοῦ αἰσθητικοῦ μέν ἐστι τὸ ἐνυπνιάξζειν, τούτου δ᾽ 1 φανταστικόν. 2. Ti & ἐστὶ τὸ ἐνύπνιον, καὶ πῶς γίνεται, ἐκ τῶν περὶ τὸν ὕπνον συμβαινόντων μάλιστ᾽ ἂν θεωρήσαιμεν. τὰ γὰρ αἰϊσθητὰ καθ᾽ ἕκαστον αἰσθητήριον ἡμῖν ἐμποιοῦσιν αἴσθησιν, καὶ τὸ γινόμενον ὕπ᾽ αὐτῶν πάθος οὐ μόνον ἐνυπάρχει ἐν τοῖς αἰσθητηρίοις ἐνεργουσῶν τῶν αἰσθήσεων, ἀλλὰ καὶ ἀπελθουσῶν.

13.

Beare gives (459a10): ‘If it were the latter [viz. an affection of the faculty of perception in the simple sense] it would be possible [when asleep] to hear and see in the simple sense’. Hett: ‘for in that case it would be possible in a dream to see and hear normally. But by understanding τὸ πάθος τοῦτο ὃ καλοῦμεν ἐνυπνίαζειν as subject of ἦν, the need for further supplementation is avoided. The present version understands the text accordingly.

15

20

25

ON DREAMS

459a

89

27. 459a8-14. Dreaming, then, is not the work of judgment. But neither is it the work of straightforward perception. Then what kind of perceiving is it?

10

45928. It is plain, then, that this affection that we call dreaming is not the work of that which judges, nor of that which engages in thought. But nor is it, | in an unqualified way, the work of that which perceives. For then it would be unqualified seeing and hearing. But just how and in what manner is what has to be examined. Let it be assumed, then, as is indeed plain, that the affection is the work of the perceptual part, granted that sleep is too. For sleeping does not belong to one part of the animal and dreaming to another. Rather, both belong to the same part. 28. 459a14-22. Conclusion: dreaming is the work of perception, but belongs to that faculty in its imagining capacity.

15

20

459214. Now since | imagination (phantasia) has been discussed in our work on the soul, and since the imagining part (to phantastikon) is the same as the perceptual, yet their being is different for the imagining (phantastikon) and the perceptual; and since imagination (phantasia) is the movement produced by a sense in the course of its active functioning, and the dream appears to be some sort of appearance (phantasma) — for an appearance (phantasma) in sleep | (whether it occurs simply or in some particular way) is what we call a dream -- it is plain that dreaming is the work of the perceptual part, but belongs to this part in its imagining (phantastikon) capacity.

Chapter Two 29. 459a23-28. To understand the nature and explanation of dreams, one should study what happens in sleep. Sense-impressions leave remnants in the sense-organs, which continue to be experienced after the original perceptions have gone. 25

459a23. What a dream is, and how it occurs, we may best study from the circumstances attending sleep. For sense-objects | corresponding to each sense-organ provide us with perception. And the affection produced by them persists in the sense-organs, not only while the perceptions are being actualized, but also after they have gone. 30. 459a28-b7.

Just as objects

can

continue

in motion,

even

when

what

started them is no longer in contact with them, so can qualitative change be

90

ΠΕΡῚ

ENYIINION

459a

παραπλήσιον γὰρ τὸ πάθος ἐπί τε τούτων καὶ ἐπὶ τῶν φερομένων ἔοικεν εἶναι. καὶ γὰρ ἐπὶ τῶν φερομένων τοῦ | κινήσαντος οὐκέτι θιγγάνοντος κινεῖται: τὸ γὰρ κινῆσαν ἐκίνησεν ἀέρα τινά, καὶ πάλιν οὗτος κινούμενος ἕτερον: καὶ τοῦτον δὴ τὸν τρόπον, ἕως ἂν στῇ, ποιεῖται τὴν κίνησιν καὶ ἐν ἀέρι καὶ ἐν τοῖς ὑγροῖς. ὁμοίως δὲ ὑπολαβεῖν τοῦτο δεῖ καὶ én’ ἀλλοιώσεως: τὸ γὰρ θερμανθὲν ὑπὸ τοῦ θερμοῦ τὸ πλησίον θερμαίνει, καὶ τοῦτο διαδίδωσιν ἕως τῆς ἀρχῆς. ὥστε καὶ ἐν τῷ αἰσθάνεσθαι, ἐπειδή ἐστιν ἀλλοίωσίς Tig ἡ κατ᾽ | ἐνέργειαν αἴσθησις, ἀνάγκη τοῦτο συμβαίνειν. διὸ τὸ πάθος ἐστὶν οὐ μόνον ἐν αἰσθανομένοις τοῖς αἰσθητηρίοις, ἀλλὰ καὶ ἐν πεπαυμένοις, καὶ ἐν βάθει καὶ

30

459b

ἐπιπολῆς. x

φανερὸν

δὲ

b

Uu

^

ὅταν

3

συνεχῶς

Pd

αἰσθανώμεθα

La

΄

TU

μεταφερόντων

γὰρ τὴν αἴσθησιν ἀκολουθεῖ τὸ πάθος, οἷον ἐκ τοῦ ἡλίου εἰς τὸ σκότος: | συμβαίνει yàp μηδὲν ὁρᾶν διὰ τὴν ἔτι ὑποῦσαν κίνησιν ἐν τοῖς ὄμμασιν ὑπὸ τοῦ φωτός. κἂν πρὸς ἕν χρῶμα πολὺν H

χρόνον ^

^

av

τὴν

΄

βλέψωμεν »

ὄψιν

^

x



»

λευκὸν

΄



#

χλωρόν, »

μεταβάλωμεν.

x

κἂν

πρὸς

^

τοιοῦτον >

t,

τὸν

,

ἥλιον

΄

ics

φαίνεται

ὅπερ

RAS

βλέψαντες

7

UV

ἐφ᾽

LA

vw



.

ἄλλο 7

2

TL λαμπρὸν μύσωμεν, παρατηρήσασι | φαίνεται κατ᾽ εὐθυωρίαν, À συμβαίνει τὴν ὄψιν ὁρᾶν, πρῶτον μὲν τοιοῦτον τὴν χρόαν, εἶτα μεταβάλλει εἰς φοινικοῦν κἄπειτα πορφυροῦν, ἕως ἂν εἰς τὴν ^

?

μέλαιναν

we

ἔλθῃ

,

χρόαν

^

^

koi

H

^

^

ἀφανισθῇ.

μεταβάλλουσιν, οἷον ἀπὸ τῶν τάχιστα | ῥεόντων, φαίνεται τὰ

^

καὶ

ποταμῶν, ἠρεμοῦντα

H

ἀπὸ

10

15

T

b

^

τῶν

L4

κινουμένων

μάλιστα δὲ κινούμενα.

ἀπὸ

*

δὲ

τῶν

γίνονται δὲ καὶ ἀπὸ τῶν μεγάλων ψόφων δύσκωφοι καὶ ἀπὸ τῶν ἰσχυρῶν ὀσμῶν δύσοσμοι, καὶ ἐπὶ τῶν ὁμοίων «ὁμοίως». ταῦτά γε δὴ φανερῶς συμβαίνει τοῦτον τὸν τρόπον.

20

4598

ON DREAMS

91

transmitted. Thus, perceptual remnants can persist when the sense-organs are no longer active. 459a28. For the affection in their case would seem akin to that of objects 30 moving in space. In their case too | there is motion, even when the moving agent is no longer in contact with them. For the moving agent moves a certain portion of air; and that, on being moved, in turn moves another. And 459b in that way motion continues to be produced, | in air and liquids alike, until a standstill is reached.

S

459b1. This, one must assume, applies in the case of alteration likewise. For what has been heated by something hot heats its neighbour, and this passes it on successively, until the starting-point is reached. So this must happen in perceiving as well, seeing that | actualized perception is a kind of alteration. Hence the affection persists in the sense-organs, both in depth and on the surface, not only while they are actually perceiving, but even after they have ceased to do so. 31. 459b7-23. Examples of perceptual after-effects in optical and other phenomena.

10

15

20

459b7. This is plain whenever we engage in perceiving something continuously. For when we shift our perception, e.g. from sunlight to darkness, our previous affection continues. | For the result is that we see nothing, because of the movement that was due to the light and is still subsisting in our eyes. Again, if we look for a long time at a single colour, be it white or green, then any object to which we may shift our vision appears to be of the same colour. And again, if we close our eyes after looking towards the sun or some other shining object, then if we watch carefully, | it appears directly in line with our original vision, first in its own colour; then it changes to crimson, next to purple, until finally it turns black and disappears. Also, when people turn away from moving objects, e.g. rivers, particularly very | fast-flowing ones, things at rest appear to them to be moving. 459b20. Then again, people are made hard of hearing by loud noises, and their sense of smell is impaired by strong smells, and likewise for similar cases. Plainly, those effects are produced in the above manner.

92

ΠΕΡῚ t

bi

oTi

δὲ

σημεῖον

ταχὺ

τὸ

x

x

λ

TA

| ἐπὶ

x

αἰσθητήρια

τῶν

459b

ἘΝΥΠΝΊΩΝ

la

^

καὶ

ἐνόπτρων

^

μικρᾶς

5

διαφορᾶς

γινόμενον:

περὶ

7

αἰσθάνεται,

οὗ

καὶ

αὐτοῦ

25

ἐπιστήσας σκέψαιτό τις ἂν καὶ ἀπορήσειεν. ἅμα δ᾽ ἐξ αὐτοῦ δῆλον ὅτι ὥσπερ καὶ ἡ ὄψις πάσχει, οὕτω καὶ ποιεῖ τι. ἐν γὰρ τοῖς ἐνόπτροις τοῖς σφόδρα καθαροῖς, ὅταν τῶν καταμηνίων ταῖς γυναιξὶ γινομένων ἐμβλέψωσιν εἰς τὸ κάτοπτρον, [γίνεται τὸ ἐπιπολῆς τοῦ ἐνόπτρου οἷον νεφέλη αἱματώδης: κἂν μὲν καινὸν N TO κάτοπτρον, οὐ ῥᾷάδιον ἐκμάξαι τὴν τοιαύτην κηλίδα, ἐὰν δὲ παλαιόν, ῥᾷον. αἴτιον δέ, ὥσπερ εἴπομεν, ὅτι οὐ μόνον πάσχει ἡ ὄψις ὑπὸ τοῦ ἀέρος, ἀλλὰ καὶ ποιεῖ τι καὶ κινεῖ, ὥσπερ καὶ τὰ λαμπρά: καὶ γὰρ ἡ ὄψις τῶν λαμπρῶν καὶ ἐχόντων χρῶμα. τὰ μὲν οὖν ὄμματα εὐλόγως, ὅταν À τὰ καταμήνια, διακεῖται ὥσπερ καὶ ἕτερον μέρος ὁτιοῦν: καὶ γὰρ φύσει τυγχάνουσι φλεβώδεις ὄντες. διὸ γινομένων τῶν καταμηνίων διὰ ταραχὴν καὶ φλεγμασίαν αἱματικὴν ἡμῖν μὲν ἡ ἐν τοῖς ὄμμασι διαφορὰ ἄδηλος, ἔνεστι δέ (ἡ γὰρ αὐτὴ φύσις σπέρματος καὶ καταμηνίων), ὁ 8' ἀὴρ κινεῖται ὑπ᾽ αὐτῶν, καὶ τὸν ἐπὶ τῶν | κατόπτρων ἀέρα συνεχῆ ὄντα ποιόν τινα ποιεῖ καὶ τοιοῦτον olov αὐτὸς πάσχει. ὁ δὲ τοῦ κατόπτρου τὴν ἐπιφάνειαν. ὥσπερ δὲ τῶν ἱματίων, τὰ μάλιστα καθαρὰ τάχιστα κηλιδοῦται: τὸ γὰρ T

3

^

x

-^

10

^

καθαρὸν ἀκριβῶς δηλοῖ 3 7 H t ἐλαχίστας κινήσεις. ὁ ^

3



^

^

τ

ὁποιασοῦν

οὖσαν x

διὰ >

Kal

πάντη. εἰσδύεται

διὰ

ἐν

^

bi



^

μὲν

τοῖς x

κηλὶς

Several mss omit distinctly whatever no reason why the corresponds to τὰ

3

ἀλλ᾽

καὶ

ὥσπερ

^

x

^

^

τῶν 3

εἰς

οὐκ

F4

~

ἔκμαξιν

t

διὰ

x

καὶ

^

βάθος,

ἐμμένει,

^

διὰ

δὲ

τὸ ^ x

δὲ

οὐχ

μὴ

^

καθαρὸν 3

καὶ

ὅτι

x

τοῦ

αἴτιον x

τρίψιν

ἀνάπλυσιν),

#

τοιούτων

15

#

οἷον

ouca.

#

^

T^

νοῆσαι

ὁπηλικηοῦν

καθαρὸν 2

x

δὲ

κατόπτρων

^

γὰρ

παλαιοῖς 5

^

(δεῖ

γίνεται

^

τὸ

x

iA

καινῶν

΄

διαίδύεται

δὲ €

τῶν

L4

μάλιστα

ἁφὴν

Y

ἐνδήηλος

ἐκ

λεῖον:

πάντῃ, 4

ἀέρος x

#

^

^

καθαρὸν ^

καὶ

M

H

Id

αἰσθάνεται

ταχέως x

εἶναι

~

τοῦ

*

τὸ

© τι ἂν δέξηται, καὶ τὸ μάλιστα τᾶς x x x x * ^ ^ δὲ χαλκὸς διὰ μὲν | τὸ λεῖος εἶναι

^

ἀφῆς

τὴν Ν

δὲ

ἀτιέναι

14.

460a

^

+

1

30

΄

εἰς

βάθος

^

λεῖον

τὸ

^

ὁμοίως

?

ἐπιπολαιότερον.

τὸ at 460al4, giving the sense ‘for anything clean shows up it receives, and especially the smallest blemishes’. But there is smallest blemishes should be especially visible, and τὸ μάλιστα μάλιστα καθαρὰ in bl2 above.

20

459b

ON DREAMS

93

32. 459b23-460a23. The sense-organs are sensitive even to slight changes and are not only affected by their objects but react upon them. This is manifest from a phenomenon on the surface of mirrors. 25

30

459b23. That the sense-organs are rapidly sensitive even to slight differences is shown by what | happens with mirrors. Indeed, anyone who would devote attention to it might look into that subject too and raise a difficulty. At the same time, it is clear from this case that the organ of sight is not only affected by, but also acts upon, its object. For in extremely clean mirrors, when

women look into them during their menstrual period, | the mirror surface

takes on a sort of blood-red cloud. In fact, if the mirror is a new one, it is not easy to get the stain out, although it is easier with an old one. The reason is, 460a as we have said, | that the organ of sight is affected not only by the air, but is also active and imparts movement, just as shining objects do. In fact the organ of sight is just such an object and one that possesses colour. One may reasonably suppose, then, that during menstrual periods the eyes are in the 5 same state as | any other part of the body. Furthermore, they are full of blood-vessels by nature. Hence, when menstruation occurs, owing to disorder and turbulence of the blood, the difference in the eyes is invisible to us, and

10

15

20

yet it is present (for the nature of semen and of the menses is the same). The air is moved by the eyes, and makes the | air extending over the mirror's surface to be of a certain quality, i.e. that by which it is affected itself. And this air in turn affects the surface of the mirror. Now just as with clothes, the cleanest are the quickest stained; for anything clean shows up distinctly whatever it receives, the most clean showing the smallest blemishes.!4 Likewise the bronze, owing to | its smoothness, is highly sensitive to any sort of impact (and one should recognise the impact of air as a form of friction, a wiping, as it were, or washing on). And because of its cleanness, any impact whatever shows up on it. The reason why the stain will not readily come off new mirrors is that the surface is clean and smooth. For it | permeates such mirrors in depth and all over — in depth because the surface is clean, and all over because it is smooth; whereas in old mirrors it does not persist, because

the stain does not penetrate to the same extent, but is more superficial.

94

ΠΕΡῚ

ἘΝΥΠΝΊΩΝ

460a

ὅτι μὲν οὖν καὶ ὑπὸ τῶν μικρῶν διαφορῶν γίνεται κίνησις, καὶ ὅτι ταχεῖα ἡ αἴσθησις, [καὶ ὅτι οὐ μόνον πάσχει, ἀλλὰ καὶ 25 ἀντιποιεῖ τὸ τῶν χρωμάτων αἰσθητήριον, φανερὸν ἐκ τούτων. μαρτυρεῖ δὲ τοῖς εἰρημένοις καὶ τὰ περὶ τοὺς οἴνους καὶ τὴν μυρεψίαν συμβαίνοντα. τό τε γὰρ παρασκευασθὲν ἔλαιον ταχέως λαμβάνει τὰς τῶν πλησίον ὀσμάς, καὶ οἱ οἶνοι τὸ αὐτὸ τοῦτο πάϊσχουσιν: οὐ γὰρ μόνον τῶν ἐμβαλλομένων ἢ ὑποκιρναμένων 30 ἀλλὰ καὶ τῶν πλησίον τοῖς ἀγγείοις τιθεμένων ἢ πεφυκότων ἀναλαμβάνουσι τὰς ὀσμάς. πρὸς δὲ τὴν τῶν εἰρημένων αἰσθητοῦ ἐμμένει Ν

x

πρὸς

ἐν

τοῖς

ἐξ ἀρχῆς | σκέψιν ὑποκείσθω ἕν φανερόν, ὅτι καὶ ἀπελθόντος τὰ αἰσθήματα αἰσθητὰ ὄντα. ^

δὲ

t^

τούτοις

πάθεσιν

ig

OTL

7

3

ῥρᾳδίως

ὄντες,

ἄλλοι

^

ἀπατώμεθα

δὲ

| ἐν

περὶ

ἄλλοις,

x

μέν, τοῦ x

5

τὰς

οἷον

ὅπερ ἐκ θύραθεν

460b

7

αἰσθήσεις

ὁ δειλὸς

ἐν

φόβῳ, ὁ δ᾽ ἐρωτικὸς ἐν ἔρωτι, ὥστε δοκεῖν ἀπὸ μικρᾶς ὁμοιότητος τὸν μὲν τοὺς πολεμίους ὁρᾶν, τὸν δὲ τὸν ἐρώμενον. καὶ ταῦτα ὅσῳ ἂν ἐμπαθέστερος À, τοσούτῳ an’ ἐλάσσονος ὁμοιότητος φαίνεται. τὸν αὐτὸν δὲ τρόπον καὶ ἐν ὀργαῖς καὶ ἐν πάσαις ἐπιθυμίαις εὐκαπάτητοι γίνονται πάντες, καὶ μᾶλλον ὅσῳ 10 ἂν μᾶλλον ἐν τοῖς πάθεσιν ὦσιν. διὸ καὶ τοῖς πυρέττουσιν ἐνίοτε φαίνεται ζῷα ἐν τοῖς τοίχοις ἀπὸ μικρᾶς ὁμοιότητος τῶν γραμμῶν συντιθεμένων. καὶ ταῦτ᾽ ἐνίοτε συνεπιτείνει τοῖς 7

Uu

πάθεσιν

οὕτως,

(^

^

OTL

Ϊ

[72

ψεῦδος,

»

ὥστε, t€

éàv

*

ἂν ^

δὲ

μέν ^

μεῖζον

μὴ T

n

^

7

,

σφόδρα x

κάμνωσι,

^

τὸ

πάθος,

ταῦτα καὶ ᾧ

τὸ τὰ

*

καὶ

μὴ

x

FA

λανθάνειν ^

κινεῖσθαι

x

πρὸς

αὐτά. αἴτιον δὲ τοῦ συμβαίνειν δύναμιν κρίνειν τό τε κύριον LA

15.

N

^

7

^

One ms (M), after ‘the former judges that he sees he sees these sticks standing nearby, he thinks elaboration of Aristotle's example is probably a phenomenon of ‘seeing X as Y', in virtue of some

^

μὴ κατὰ τὴν αὐτὴν φαντάσματα γίνεται. ^

\

b

b

bi

his enemies’ (460b7), adds 'e.g., if he is seeing armed men’. This gloss. But it illustrates aptly the slight resemblance between them.

460a

ON DREAMS

95

33. 460a23-32. The same conclusions are evidenced by wines and ointments. 25

30

460223. From the above, then, it is plain that movement is generated even by slight differences; that perception is rapid; and | that the sense-organ that perceives colours is not only affected by them but also acts reciprocally. What has been said is further attested by what happens with wines and the preparation of ointments. For when oil has been prepared, it rapidly takes on the smells of things near it. Wines too are affected in the same way: | they pick up the smells not only of things thrown or mixed in with them, but also of things placed near their containers or growing nearby. 34. 460a32-b16. Original inquiry resumed. impressions persist after external stimuli have liable to perceptual error when emotionally varies according to our condition and may judgment.

Two assumptions: (1) Senseceased. (2) We are especially excited. The degree of error or may not be endorsed by

460b 460a32. Returning, then, to our original ! inquiry, let one thing be assumed, that is plain from what has been said: our sense-impressions persist, remaining perceptible, even after the external sense-object has gone.

5

10

15

460b3. It may be added that we are easily deceived with respect to our perceptions while we are in emotional states. And different people laccording to different states, e.g. the coward in a state of fright, the amorous man in one of amorous passion. Thus, from a slight resemblance the former judges that he sees his enemies, ? but the latter that he sees his loved one. The more emotional his state, the slighter the resemblance that can give rise to these appearances. In the same way, all men become prone to | deception while in states of anger as well as in every form of appetite, and the more so, the more they are in those states. That is also why animals sometimes appear on the walls to people in a fever, from a slight resemblance in the combination of lines. Sometimes, also, those appearances correspond to their state in such a way that if they are not seriously ill, they are aware of the | illusion; whereas if their condition is more serious, they actually move themselves in accordance with the appearances. 35. 460b16—27. judgment.

Explanation

of the

conflict

between

appearance

and

460b16. The reason why these things happen 15 that the ruling part and that by which appearances (phantasmata) occur do not judge on the basis of the

96

ΠΕΡῚ

ENYHNION

460b

τούτου δὲ σημεῖον ὅτι φαίνεται μὲν ὁ ἥλιος ποδιαῖος, ἀντίφησι δὲ πολλάκις ἕτερόν τι πρὸς τὴν φαντασίαν. καὶ τῇ ἐπαλλάξει τῶν δακτύλων τὸ Ev δύο φαίνεται, ἀλλ᾽ ὅμως οὔ φαμεν δύοκυριωτέρα γὰρ τῆς ἀφῆς ἡ ὄψις. εἰ & ἦν ἡ doh μόνη, κἂν ἐκρίνομεν τὸ ἕν δύο. τοῦ δὲ διεψεῦσθαι αἴτιον ὅτι οὐ μόνον τοῦ αἰσθητοῦ κινοῦντος φαίνεται ἀἁδήποτε, ἀλλὰ καὶ τῆς αἰσθήσεως κινουμένης αὐτῆς, ἐὰν ὡσαύτως κινῆται ὥσπερ καὶ ὑπὸ τοῦ αἰσθητοῦ: λέγω δ᾽ οἷον ἡ γῆ δοκεῖ τοῖς πλέουσι κινεῖσθαι κινουμένης τῆς ὄψεως ὑπ᾽ ἄλλου. 3. Ἐκ δὴ τούτων φανερὸν ὅτι οὐ μόνον ἐγρηγορότων αἱ κινήσεις αἱ ἀπὸ τῶν αἰσθημάτων γινόμεναι τῶν τε θύραθεν | καὶ τῶν ἐκ τοῦ σώματος ἐνυπάρχουσιν, ἀλλὰ καὶ ὅταν γένηται τὸ πάθος τοῦτο ὃ καλεῖται ὕπνος, καὶ μᾶλλον τότε φαίνονται. μεθ᾽ ἡμέραν μὲν γὰρ ἐκκρούονται ἐνεργουσῶν τῶν [αἰσθήσεων καὶ τῆς διανοίας, καὶ ἀφανίζονται ὥσπερ παρὰ πολὺ πῦρ ἔλαττον καὶ λῦπαι καὶ ἡδοναὶ μικραὶ παρὰ μεγάλας, παυσαμένων δὲ ἐπιπολάζει καὶ τὰ μικρά: νύκτωρ δὲ δι᾽ ἀργίαν τῶν κατὰ μόριον αἰσθήσεων καὶ ἀδυϊναμίαν τοῦ ἐνεργεῖν, διὰ τὸ ἐκ τῶν ἔξω εἰς τὸ ἐντὸς γίνεσθαι τὴν τοῦ θερμοῦ παλίρροιαν, ἐπὶ τὴν ἀρχὴν τῆς αἰσθήσεως καταφέρονται καὶ γίνονται φανεραὶ καθισταμένης τῆς

δὲ

ποταμοῖς M

ὑπολαβεῖν +

σχήματα

πάμπαν

πολλὴ

yàp

x

νέοις



μὲν ^

διὰ

τὴν

οὖσιν,

κίνησις

τὰς

οὕτω x

πολλάκις [4

ἄλλα

ὥσπερ

γινομένας,

^

συνεχῶς,

16.

25

30

4618

ταραχῆς. δεῖ

καὶ

20

διὰ

τὴν

f

F4

ὁμοίως, 5

μικρὰς

δίνας

κίνησιν 7

πολλᾶκις

LA

\

ἀντίκρουσιν.

διὸ

παιδίοις,

ἀπὸ

τῆς ^

γίνεσθαι Ζ

^

τοῖς

οὐ

τροφῆς

3

διαλυομένας μετὰ

^

τοῖς

ἑκάστην

καὶ

τὴν

ἐν

s

δὲ

οἷον

τὰς

\

^

^

τὴν

γίνεται

εἰς

τροφὴν

ἐνύτινια"

θερμότητα.

ὥστε t^

Alternatively, reading κρίνειν τὸ κύριον καὶ τὰ φαντάσματα γίγνεσθαι with EM (460b17-18), "The reason why these things happen is that it is not on the basis of the same faculty that the ruling part judges and the appearances occur. This avoids the implication of Ross's text that 'the part by which appearances occur makes a judgment.

10

460b 20

25

ON DREAMS

97

same faculty.!© Proof of this is the fact that the sun appears only one foot across, and yet frequently something else contradicts the appearance (phantasia). | Again, by crossing of the fingers a single object appears as two, but even so we still deny that there are two things. For sight has more authority than touch. If touch stood alone, we should actually judge the single object to be two. 460b22. The reason for the illusion is that any kind of appearance may arise, not only when a sense-object is imparting the movement, but also when a sense is itself being moved, | provided it be moved just as it would be by the sense-object. I mean, for example, that to people sailing past, the land seems to be moving, when it is vision that is being moved by something else. Chapter Three 36. 460b28-461a8. Perceptual movements, then, can affect not only waking people, but sleeping ones as well. Persisting movements are especially manifest during sleep, because normal perception and thought are then suspended. 460b28. From the above it is plain that the movements arising from sense-

30

impressions,

both those coming

from

outside

| and those from within

the

body, are present not only when people are awake, but also whenever the affection called sleep comes upon them, and that they are especially apparent 461a at that time. For in the day-time, while the | senses ard the intellect are functioning, they are pushed aside or obscured, like a smaller fire next to a large one, or minor pains and pleasures next to big ones, though when the latter cease, even the minor ones come to the surface. By night, however, 5 owing to the inactivity of the special senses and their ! inability to function, because of the reversed flow of heat from the outer parts to the interior, they are carried inward to the starting-point of perception, and become apparent as the disturbance subsides. 37. 461a8-25. Remnants of perception may be obliterated or may appear during sleep in a distorted form. 10

461a8. One must assume that like small eddies forming in rivers, similarly each movement develops | continuously, often in a uniform pattern, but often breaking up into different shapes because of obstruction. That is why dreams do not occur after food or in the very young, e.g. infants. For there is much movement because of the heat generated from the food. Hence, just as in a

98

ΠΕΡῚ

Καθάπερ εἴδωλον, φαίνεσθαι φανερά, ὑπόλοιποι μὲν ὑπὸ

ENYIINION

461a

ἐν ὑγρῷ, ἐὰν σφοΐδρα κινῇ τις, ὁτὲ μὲν οὐθὲν φαίνεται ὁτὲ δὲ φαίνεται μέν, διεστραμμένον δὲ πάμπαν, ὥστε ἀλλοῖον ἢ οἷόν ἐστιν, ἠρεμήσαντος δὲ καθαρὰ καὶ οὕτω καὶ ἐν τῷ καθεύδειν τὰ φαντάσματα καὶ αἱ κινήσεις αἱ συμβαίνουσαι ἀπὸ τῶν αἰσθημάτων ὁτὲ | μείζονος οὔσης τῆς εἰρημένης κινήσεως ἀφανίζονται

πάμπαν,

ὁτὲ

δὲ

τεταραγμέναι

φαίνονται

τερατώδεις, καὶ οὐκ εἰρόμενα τὰ ἐνύπνια, κοῖς καὶ πυρέττουσι καὶ οἰνωμένοις: πάντα πνευματώδη ὄντα πολλὴν ποιεῖ κίνησιν καὶ

αἱ

ὄψεις

3

7

καὶ

κινουμένην,

ὁρᾶν

3

x

΄

μὲν

ἀλλ᾽ ἂν κίνησιν.

τὸ

ὥσπερ

e

τὸ

^



οὖν

δύο

ἀρχή,

πάντως,

ἐπικρῖνον

δ᾽ εἴπομεν

ὅτι

wv

τὴν n

Ev

φησιν

φαίνεται

^

TO

φαμεν,

x

εἰσαγγέλλειν

αἰσθήσεώς

x

διὰ

,

ὄψιν

οἷον τοῖς μελαγχολιγὰρ τὰ τοιαῦτα πάθη ταραϊχήν.

καὶ

τῷ a

ἐὰν

ἁφὴν \

ὅλως

μὴ

δοκεῖ

οὐ



25

| κυριωτέρα

ἀντιφῆ.

P4

ὁμοιότητα

17.

*

x

φαίνεται

(14

δι᾽ ἄλλο

ἐκεῖνο.

^

ὅταν

\

τὸ

κινῆται

πάθος ^

γὰρ

ἀφ᾽

t

φαινόμενον,

τὴν

οἰκείαν

καθεύδῃ,

οὕτως

αἰσθητήρια καὶ x ^ Y τὸ μικρὰν ἔχον 3

461b +

΄

εὐαπάτητοι, x

30

κινήσεις ἑκάστης

τὸ

ὁ καθεύδων διὰ τὸν ὕπνον καὶ τὸ κινεῖσθαι τὰ *7 x ΄ ^ x v t τἄλλα τὰ συμβαίνοντα | περὶ τὴν αἴσθησιν, dore €

3

οὐ

5

πάντως

μὴ

^

δοκεῖν,

δύο \

γὰρ

ἑτέρα

δὲ

κατέχηται

^

κινεῖσθαι

τὴν t^

δοκεῖ.

ἄλλοι

#

ἐνίοτε

20

καὶ

καθισταμένου δὲ καὶ διακρινομένου τοῦ αἵματος ἐν τοῖς ἐναίμοις, σῳζομένη τῶν αἰσθημάτων ἡ κίνησις ἀφ᾽ ἑκάστου τῶν αἰσθητηρίων εἰρόμενά τε ποιεῖ τὰ ἐνύπνια, καὶ φαίνεσθαί τι καὶ δοκεῖν διὰ μὲν τὰ ἀπὸ τῆς ὄψεως καταφερόμενα ὁρᾶν, διὰ δὲ τὰ ἀπὸ τῆς ἀκοῆς ἀκούειν, ὁμοιοτρόπως | δὲ καὶ ἀπὸ τῶν ἄλλων αἰσθητηρίων: τῷ μὲν γὰρ ἐκεῖθεν ἀφικνεῖσθαι τὴν κίνησιν πρὸς τὴν ἀρχὴν καὶ ἐγρηγορὼς δοκεῖ ὁρᾶν καὶ ἀκούειν καὶ αἰσθάνεσθαι,

15

LA

κατιόντος

^

τοῦ

'So that the dreams are not coherent’. ‘Coherent’ translates εἰρόμενα (461422), Drossaart Lulofs' emendation (1947, 1xii-Ixiv) for various ms readings that do not give satisfactory sense here and in line 27 below. At Rhet. 1409a24-34 Aristotle uses εἰρομένη λέξις for a ‘continuous’ form of discourse. That kind of 'coherence' is somewhat remote from the sense required here, but the notion of 'coherence', suits the context better than any alternative. Moreover, Aristotle uses the cognate verb συνείρουσιν at 464b4. Keeping éppopéva, the most plausible of the ms readings, we must translate 'the dreams are morbid' (Hett), 'the dreams are unhealthy' (Beare).

10

4618

15

ON DREAMS

99

liquid, if one | agitates it violently, sometimes no image appears, whereas at other times one does appear but so completely distorted as to appear different from what it is like, and yet when the motion has ceased, the images are clear

20

and plain; so also during sleep, the appearances (phantasmata) and residual movements deriving from sense-impressions are sometimes completely obscured by | the aforesaid movement, when it is too great; but at other times the vision appears disturbed and grotesque, so that the dreams are not coherent," as with those who are atrabilious, feverish, or intoxicated. For all such conditions, being gaseous, produce much commotion and turbulence. 38. 461a25-b7. Or sense-remnants may retain their original form, causing an appearance to occur, so that the sleeper thinks that he is perceiving something.

25

461a25. When in sanguineous animals the blood has subsided and its purer elements have separated off, the movement of sense-impressions persisting from each of the sense-organs makes the dreams coherent. Thus something is made to appear, and because of effects carried inward from vision one judges that one is seeing, or because of those from hearing, that one is hearing; and 30 soon similarly | for those from the other senses. For even when one is awake, it is because the movement from those sources reaches the starting-point that 461b one judges | that one is seeing, hearing, or perceiving. And it is because vision is sometimes judged to be moved, when it is not, that we say we are seeing; and because touch reports two movements, that one object is judged to be two. For in general the starting-point affirms the report from each sense, 5 provided that some other, | more authoritative one does not contradict it. In every case, then, something appears, yet what appears is not in every case judged to be real; it is, though, if the critical part is held in check or fails to move with its own proper movement. 39. 461b7-30. During sleep judgment is disabled from exercising its function. Thus, a perceptual remnant bearing some resemblance to a senseimpression is mistaken for the real one.

10

461b7. Now, just as we said that different people are prone to deception on account of different emotional states, so is the sleeping person on account of sleep, because his sense-organs are being moved and because of other circumstances | attending perception. Consequently, what bears a slight resemblance to something appears to be that very thing. For whenever one is asleep, as most of the blood sinks down to the starting-point, the movements

100

ΠΕΡῚ

ENYIINION

461b

πλείστου αἵματος ἐπὶ τὴν ἀρχὴν συγκατέρχονται αἱ ἐνοῦσαι κινήσεις, αἱ μὲν δυνάμει αἱ δὲ ἐνεργείᾳ. οὕτω δ᾽ ἔχουσιν ὥστε ἐν τῇ κινήσει THE ἥδε ἐπιπολάσει ἐξ αὐτοῦ ἡ κίνησις, ἂν δ᾽ αὕτη φθαρῇ, | ἥδε. καὶ πρὸς ἀλλήλας δὴ ἔχουσιν ὥσπερ οἱ πεπλασμένοι βάτραχοι οἱ ἀνιόντες ἐν τῷ ὕδατι τηκομένου τοῦ ἁλός — οὕτως ἔνεισι δυνάμει, ἀνειμένου δὲ τοῦ κωλύοντος ἐνεργοῦσιν, καὶ λυόμεναι ἐν ὀλίγῳ τῷ λοιπῷ αἵματι τῷ ἐν τοῖς αἰσθητηρίοις κινοῦνται, ἔχουσαι ὁμοιότητα ὥσπερ τὰ ἐν τοῖς νέφεσιν, ἃ παρεικάζουσιν ἀνθρώποις καὶ κενταύροις ταχέως μεταβάλλοντα. τούτων δὲ ἕκαστόν ἐστιν, ὥσπερ εἴρηται, ὑπόλειμμα τοῦ ἐν τῇ ἐνεργείᾳ αἰσθήματος: καὶ ἀπελθόντος τοῦ ἀληθοῦς ἔνεστι, καὶ ἀληθὲς εἰπεῖν ὅτι τοιοῦτον οἷον Κορίσκος, ἀλλ᾽ ,

οὐ

3

κύριον

Κορίσκος.

,

καὶ

ἀληθινὸν. J

ὅτε

,

τὸ ὃ

,

es

tr

ἐπικρῖνον,

δὴ

x

καὶ

^

δὲ

ἠσθάνετο,

x

*

ἀλλὰ

διὰ

αἰσθανόμενον δ

οὐκ

^

Κορίσκον

ἐὰν

^

τὸ

^

ἐκεῖνον

τοῦτο,

LA

Koptloxov

LA

τοῦτο

λέγει

,

ἔλεγε

+

μὴ

ἮΝ

^

15

20

x

25

τὸν

παντελῶς ^

κατέχηται ὑπὸ τοῦ αἵματος, ὥσπερ αἰσθανόμενον τοῦτο κινεῖται ὑπὸ τῶν κινήσεων τῶν ἐν τοῖς αἰσθητηρίοις, καὶ δοκεῖ τὸ ὅμοιον αὐτὸ εἶναι τὸ ἀληθές- καὶ τοσαύτη τοῦ ὕπνου ἡ δύναμις ὥστε ποιεῖν

τοῦτο ^

ὥσπερ

ὀφθαλμῷ, δὲ

μὴ

οὖν

οὐ

εἴ

τινα

μόνον

λανθάνῃ,

λανθάνοι

φανεῖται

φανεῖται

ὑποβαλλόμενος

ἀλλὰ μὲν

οὐ

καὶ

δόξει

δόξει

18.

19.

ἮΝ

^

h

A

t

,



εἶναι

δέ,

ὕπνοις, ἐὰν μὲν αἰσθάνηται ὅτι καθεύδει, καὶ αἴσθησις τοῦ ὑπνωτικοῦ, φαίνεται μέν, λέγει (74

30

λανθάνειν.

Ν

^

δάκτυλος

δύο

οὕτω

καὶ

τὸ

τῷ

ἕν, ἐν

ἂν

462a

τοῖς

τοῦ πάθους ἐν ᾧ ἡ δέ τι ἐν αὐτῷ ὅτι ^

FA

3

KA

Or, reading ἐάν τι κινήσῇ τὸ αἷμα at 461013--14 with EM, ‘They are so related that, if anything moves the blood, some one sensory movement will emerge from it’ (Beare). This enables ἐξ αὐτοῦ to refer to the blood rather than to its movement or (Ross, 277) to the fact that it is travelling to the heart. Hett gives (461b24): ‘it is true to say that it is (for instance) like Coriscus'. But οἷον should be taken simply as correlative to τοιοῦτον, not as meaning ‘for instance’. Beare gives ‘it is correct to say of it, that though not actually Koriskos, it is like Koriskos'. But it is not clear whether ἀλλ᾽ οὐ Κορίσκος should be taken as part of the ὁτι clause, i.e. as part of what it is true to say, or as an independent main clause. Alternative translations would be: 'it is true to say that it is like Coriscus; nevertheless it is not Coriscus’, or ‘it is true to say that it is like Coriscus, but (it is) not (true to say) that it is Coriscus'.

f

4610

15

ON DREAMS

101

present within it — some potentially, but some actually ~ go down with it. They are so disposed that in any given movement of the blood, one movement will rise from it to the surface;!® and if that one perishes, then | another will do so. In fact, relative to one another, they are just like those artificial frogs that float upwards in water as the salt dissolves — just so, the movements are there potentially, but become activated as soon as what impedes them is removed. Upon being released, they move in the little blood

remaining in the sense-organs, while taking on a resemblance, as cloud20

25

30

formations do, which | people liken now to men and now to centaurs as they change rapidly. Each of these, as has been said, is a remnant of the actual sense-impression, and is still present within, even when the real one has departed. Thus, it is true to say that it is like Coriscus, even though it is not Coriscus.!? While one was perceiving, | one's ruling and judging part was saying not that the sense-impression is Coriscus, but because of that impression, that the actual person out there is Coriscus. The part that says this while it is actually perceiving (unless it is completely held in check by the blood) is moved by the movements in the sense-organs, as if it were perceiving.? Consequently, what is like something is judged to be that very thing. And | the effect of sleep is sufficient to make this escape notice. 40. 461b30-462a8. Moreover, as with other illusions, the dream appearance

is endorsed by judgment, asleep. 4628

unless the dreamer should be aware

of being

461b30. Accordingly, just as for someone who was unaware of a finger being pressed beneath his eye, | a single object would not only appear two, but would actually be judged to be two, whereas for someone aware of it, it will appear but will not be judged to be two; so during periods of sleep: if someone perceives that he is asleep, i.e. that it 15 a sleeping state in which the perception is occurring, then there is an appearance, but something in him

20.

Itis unfortunate that the text of this crucial passage (461b26—29) is uncertain. Some mss give the relative pronoun in b26 as οὗ and others as @. It is difficult to make sense of either reading.

Ross (278) takes both to be corrupt, emends

the pronoun to

6, and takes its antecedent to be ‘the ruling and judging part (025), referred to again by τοῦτο in b28. He translates 'this faculty, as though it were perceiving, is affected by the movements in the sense-organs, and that which is like (the image which looks like Coriscus) seems to it to be the real object’. Ross's text has been adopted, and his interpretation followed in the present translation and in the notes on b21-30 (p.

151).

102

ΠΕΡῚ

ENYIINION

φαίνεται μὲν Κορίσκος, οὐκ καθεύδοντος λέγει τι ἐν τῇ ἽΝ

x

eav

,

δὲ

t,

λανθάνῃ

[zd

b

OTL

3

δὲ

ἀληθῆ

^

ἔστι δὲ ὁ Κορίσκος (πολλάκις yàp ψυχῇ ὅτι ἐνύπνιον τὸ φαινόμενον):

^

ὅτι

4

καθεύδει, P4

M

>

οὐδὲν ^

λέγομεν

462a

3

καὶ

,

^

ἀντιφήσει

^

7

εἰσὶ

κινήσεις

,

τῇ

φαντασίᾳ.

φανταστικαὶ

x

2

^

ἐν

τοῖς

αἰσθητηρίοις, δῆλον, ἐάν τις προσέχων πειρᾶϊται μνημονεύειν ἃ πάσχομεν καταφερόμενοί τε καὶ ἐγειρόμενοι: ἐνίοτε γὰρ τὰ φαινόμενα εἴδωλα καθεύδοντι φωράσει ἐγειρόμενος κινήσεις οὔσας ἐν τοῖς αἰσθητηρίοις: καὶ ἐνίοις γε τῶν νεωτέρων καὶ ΄

£

πάμπαν

t

κινούμενα, ἐκ

ἐνύπνιον Y

x

μὲν

γὰρ

φωτὸς $^

Ὥδη

*

x

ἐν

΄

φαίνεται:

οὐδὲ

,ὔ

^

)

τῷ

τὰ

ὕπνῳ

ὅτι

ἄρτι

ἀσθενικῶς #

£e

ὑποβλέποντες,

b



x

λεχθέντα

πᾶν. x

mm

μέντοι

τὸ

Ν

^

λελυμένων

φάντασμα F4

ἐστὶ 5

"

ἄλλο

#

15 Ὁ

γὰρ

Y

TL

αἰσθάνεϊσθαί

Lj

καθεύδειν

Y

Σ

καὶ

ἀφῆς,

ὕπνῳ: εἴ

ἐν ^

συμβαίνει

καὶ

Los

οὐδ᾽

x

πολλὰ

φοβουμένους. id

ἐν

τὸ

7

εἴδωλα

n

συλλογίσασθαι

^

καὶ

ἐνύπνια,

χυμοῦ 2

γὰρ

1

τι

ἐστιν

ἐνίοις

καὶ

μέν

Y

δεῖ

^

Y

φαίνεται

πολλαΐκις

ἁπάντων τ

,

σκότος,

ἐγκαλύπτεσθαι

φάντασμα

αἰσθήσεων

,



"n

,

οὐκ

^

ἐᾶν

2

τούτων

N

1

ElL\OWAG

5

ὥστ᾽

δὴ

>

Ν

l4

διαβλέπουσιν,

n

[10

καὶ

καὶ LA

t

πρῶτον 7

x

ψόφων

οἷον

ἠρέμα

τῶν καὶ

20

πόρρωθεν"

Pd

€opov

^

φῶς

τοῦ

^

λύχνου καθεύδοντες, ὡς ᾧοντο, ἐπεγερθέντες εὐθὺς ἐγνώρισαν τὸ τοῦ λύχνου ὄν, καὶ ἀλεκτρυόνων καὶ κυνῶν φωνὴν ἠρέμα ἀκούοντες ἐγερίθέντες σαφῶς ἐγνώρισαν. ἔνιοι δὲ καὶ ἀποκρίνονται ἐρωτώμενοι: ἐνδέχεται γὰρ τοῦ ἐγρηγορέναι καὶ καθεύδειν ἁπλῶς θατέρου ὑπάρχοντος θάτερόν πῃ ὑπάρχειν. ὧν οὐθὲν ἐνύπνιον φατέον, οὐδ᾽ ὅσαι δὴ ἐν τῷ ὕπνῳ γίνονται ^

ἀληθεῖς >

^

^

EVVOLAL vw

f

παρὰ ^

τῆς | κινήσεως τῶν καθεύδει, τοῦτ᾽ ἐστὶν

x

τὰ

x

^T

φαντάσματα, n

25

Á

αἰσθημάτων, ἐνύπνιον.

ἀλλὰ 5

ὅταν

^

ἐν

τὸ

φάντασμα

M

τῷ

LA

^

καθεύδειν

τὸ

ἀπὸ

x

3

=

1,

Ν

4

à

30

462a 5

-

ON DREAMS

103

says | that it appears to be Coriscus, and yet Coriscus is not there. (For often something in the soul of a person asleep says that what is appearing is a dream). But if he is unaware that he is asleep, then nothing will contradict the

appearance (phantasia). 41. 462a8-15. Experiences that some people have when falling asleep or waking up support the foregoing account. 462a8. That our account is true, and that there are apparitional movements (kineseis

10

15

phantastikai)

in the sense-organs,

will be clear if one makes

an

effort | to remember how we are affected when dropping off to sleep or when waking up. For, when waking up, one will occasionally detect the images appearing while one was asleep as movements in the sense-organs. Indeed, to some youngsters, even when their eyes are wide open, if it is dark, moving images frequently appear, so that they often cover their heads | in fright. 42. 462a15-31. These, however, and other experiences during sleep, are not dreams proper. Conclusion: a dream, properly speaking, is an appearance resulting from perceptual remnants, that occurs during sleep and in virtue of being asleep.

20

25

30

462215. From all of this one must conclude that a dream is a certain sort of appearance (phantasma), and specifically one that occurs during sleep. For the images just mentioned are not dreams, nor is any other that may appear once the senses have been set free. Nor again is every appearance (phantasma) during sleep a dream. For, in the first place, some people have the experience of actually perceiving, | in a certain manner, sounds and light, flavour and touch, albeit faintly and as if from a distance. For it has been known for people with their eyes partly open during sleep to recognize, as soon as they wake up, that what they saw dimly while asleep, as the light of a lamp (as they supposed) was indeed from the lamp. Or again, after faintly hearing the noise of cocks and dogs, | they recognized these clearly upon waking up. Again, some people actually give answers when questioned. For in the case of waking and sleeping, when one of these states is present in the ordinary way, it is possible for the other to be present in a certain manner. None of those things should be called a dream. Nor should any true thoughts that may occur during sleep, over and above the appearances (phantasmata). Rather, it is an appearance (phantasma) that arises from the | movement of the sense-impressions, while one is in the sleeping state and in virtue of one's being asleep, that is the dream proper.

104

ΠΕΡῚ

ἘΝΥΠΝΊΩΝ

4620

ἤδη δέ τισι συμβέβηκεν μηϊδὲν ἐνύπνιον ἑωρακέναι κατὰ τὸν βίον, τοῖς δὲ πόρρω mou προελθούσης τῆς ἡλικίας ἰδεῖν πρότερον μὴ ἑωρακόσιν. τὸ δ᾽ αἴτιον τοῦ μὴ γίνεσθαι παραπλήσιον φαίνεται τῷ ἐπὶ τῶν παιδίων καὶ μετὰ τὴν τροφήν. ὅσοις γὰρ συνέστηκεν ἡ φύσις ὥστε πολλὴν ἀναθυμίασιν πρὸς τὸν ἄνω τόπον ἀναφέρεσθαι, ἣ πάλιν καταφερομένη ποιεῖ πλῆθος κινήσεως, εὐλόγως τούτοις οὐδὲν φαίνεται φάντασμα. προϊούσης δὲ τῆς ἡλικίας οὐδὲν ἄτοπον φανῆναι ἐνύπνιον: μεταβολῆς γάρ τινος γενομένης ἤ καθ᾽ ἡλικίαν ἤ κατὰ πάθος ἀναγκαῖον συνβῆναι τὴν ἐναντίωσιν ταύτην.

21.

A variant ms tradition, after κατὰ τοιοῦτόν

ἐστι,

συμβαΐνει

δ᾽

τὸν

ὅμως.

βίον, (4620), adds σπάνιον καὶ

τοῖς

μὲν

ὅλως

μὲν

διετέλεσεν:

οὖν

τὸ

‘sucha

condition is rare, but it occurs nonetheless. For some it persists throughout life . . .. There are further minor variants from ὅσοις (line 5) onwards, but they do not affect the sense.

4620

5

10

4620

ON DREAMS

105

43. 462a31-b11. Addendum regarding non-dreamers. 462b 462a31. There have been cases of people who have not experienced | a single dream in the course of a life-time, or who have experienced one when

5

10

well

advanced in years, without having done so before.?! The reason for the nonoccurrence appears | akin to that which operates in the case of infants and after food. Thus, it is intelligible that no appearance (phantasma) should appear to those so constituted by nature that much exhalation 1s carried up to the upper region, which produces an abundance of motion upon returning downward. As their age advances, however, it is not surprising that a dream

should appear. | For when a certain change has developed, according either to age or to condition, this reversal must necessarily take place.

106

462b

ΠΕΡῚ

ΤῊΣ

KAO’

YIINON

MANTIKHX

1. Περὶ δὲ τῆς μαντικῆς τῆς ἐν τοῖς ὕπνοις γινομένης καὶ λεγομένης συμβαίνειν ἀπὸ τῶν ἐνυπνίων, οὔτε καταφρονῆσαι ῥάδιον οὔτε πεισθῆναι. τὸ μὲν γὰρ πάντας ἢ πολλοὺς ὑπολαμβάνειν ἔχειν τι σημειῶδες τὰ ἐνύπνια παρέχεται πίστιν ὡς ἐξ ἐμπειρίας λεγόμενον, καὶ τὸ περὶ ἐνίων εἶναι τὴν μαντικὴν ἐν τοῖς ἐνυπνίοις οὐκ ἄπιστον: ἔχει γάρ τινα λόγον: διὸ καὶ περὶ τῶν ἄλλων ἐνυπνίων ὁμοίως ἄν τις οἰηθείη. τὸ δὲ μηδεμίαν αἰτίαν εὔλογον ὁρᾶν καθ’ ἣν ἂν γίνοιτο, τοῦτο | δὴ ἀπιστεῖν ποιεῖ: τό τε γὰρ θεὸν εἶναι τὸν πέμποντα, πρὸς τῆ ἄλλῃ. ἀλογίᾳ, καὶ τὸ μὴ τοῖς βελτίστοις καὶ φρονιμωτάτοις ἀλλὰ τοῖς τυχοῦσι πέμπειν ἄτοπον. ἀφαιρεθείσης δὲ τῆς ἀπὸ τοῦ θεοῦ αἰτίας οὐδεμία τῶν ἄλλων εὔλογος εἶναι φαίνεται αἰτία: τοῦ γὰρ περὶ τῶν ἐφ᾽ Ἡρακλείαις στήλαις | ἢ τῶν ἐν Βορυσθένει προορᾶν twas ὑπὲρ τὴν ἡμετέραν εἶναι δόξειεν ἂν σύνεσιν εὑρεῖν τὴν

15

20

25

ἀρχήν. ἀνάγκη δ᾽ οὖν τὰ ἐνύπνια ἢ αἴτια εἶναι ἢ σημεῖα τῶν γινομένων ἢ συμπτώματα, ἢ πάντα ἢ ἔνια τούτων ἢ Ev μόνον. λέγω δ᾽ αἴτιον μὲν οἷον τὴν σελήνην τοῦ ἐκλείπειν τὸν ἥλιον, καὶ τὸν κόπον | τοῦ πυρετοῦ, σημεῖον δὲ τῆς ἐκλείψεως τὸ τὸν ἀστέρα εἰσελθεῖν, τὴν δὲ τραχύτητα τῆς γλώττης τοῦ πυρέττειν, σύμπτωμα δὲ τὸ βαδίζοντος ἐκλείπειν τὸν ἥλιον: οὔτε yàp ow | μεῖον τοῦ ἐκλείπειν τοῦτ᾽ ἐστὶν οὔτ᾽ αἴτιον, οὔθ᾽ ἡ ἔκλειψις τοῦ βαδίζειν: διὸ τῶν συμπτωμάτων οὐδὲν οὔτε ἀεὶ γίνεται, οὔθ᾽ ὡς EM τὸ πολύ. τῶν

ap’ οὖν ἐστι τῶν ἐνυπνίων τὰ μὲν αἴτια, τὰ δὲ σημεῖα, οἷον περὶ τὸ σῶμα συμβαινόντων; λέγουσι | γοῦν καὶ τῶν ἰατρῶν

30

463a

4620

107

ON DIVINATION THROUGH

SLEEP

Chapter One 44. 462b12-26. Considerations for and against belief in divination from dreams.

15

20

25

462b12. As for the divination that takes place during periods of sleep and is said to be based on dreams, it is not easy either to despise it or to believe in it. The fact that all or many people ! suppose dreams to have some significance inspires belief in it, as deriving from experience. And as regards some matters it is not beyond belief that there should be divination during dreams, because it does make some sense. One might, accordingly, take a similar view of other dreams as well. And yet the fact of seeing no reasonable explanation for its occurrence | makes for disbelief. For, apart from its general irrationality, the idea that it is God who sends dreams, and yet that he sends them not to the best and most intelligent, but to random people, is absurd. Yet if the explanation ascribing them to God is ruled out, none of the others appears to be reasonable. For the source of certain people's prevision of events at the Pillars of Heracles | or on the Borysthenes would seem beyond our wit to discover. 45. 462b26-463a3. The dreams in question must be either causes or signs of the events they portend, or else coincidences, one or more of these.

462b26. Well then, it is necessary that the dreams are either causes or signs of things that happen, or else coincidences; either all or some of these, or one only. By a cause, I mean, for example, the moon as a cause of the sun's being 30 eclipsed, or fatigue | as a cause of a fever. By a sign, the star's entry into shadow as a sign of its eclipse, or roughness of the tongue as a sign of someone's having a fever. And by coincidence, the sun's being eclipsed 463a when someone is taking a walk, since that is neither a | sign nor a cause of its being eclipsed, nor is the eclipse of the walking. Hence no coincidence happens either always or for the most part. 46. 463a3—21. They may be signs. For they may presage illnesses or other imminent bodily conditions. Early intimations of these would be more noticeable in sleep. 46323. Is it true, then, that some dreams are causes, while others are signs,

5

e.g. of what is happening about the body?

At | all events, even medical

108

ΠΕΡῚ

ΤῊΣ

ΚΑΘ᾽

YIINON

463a

MANTIKHZ

οἱ χαρίεντες ὅτι δεῖ σφόδρα προσέχειν τοῖς ἐνυπνίοις. εὔλογον δὲ οὕτως ὑπολαβεῖν καὶ τοῖς μὴ τεχνίταις μέν, σκοπουμένοις δέ τι καὶ φιλοσοφοῦσιν. at γὰρ μεθ᾽ ἡμέραν γινόμεναι κινήσεις, ἂν μὴ σφόδρα μεγάλαι ὦσι καὶ ἰσχυραί, λανθάνουσι παρὰ μείζους τὰς Eypnyopikäs κινήσεις, ἐν δὲ τῷ καθεύδειν τοὐναντίον: καὶ γὰρ αἱ μικραὶ μεγάλαι δοκοῦσιν εἶναι. δῆλον 8 ἐπὶ τῶν συμβαινόντων κατὰ τοὺς ὕπνους πολλάκις: οἴονται γὰρ κεραυνοῦσθαι καὶ βροντᾶσθαι μικρῶν ἤχων ἐν τοῖς ὠσὶ γινομένων, καὶ μέλιτος καὶ γλυκέων χυμῶν ἀπολαύειν ἀκαριαίου φλέγματος καταρρέοντος, καὶ βαδίζειν διὰ πυρὸς καὶ θερμαίνεσθαι σφόδρα μικρᾶς θερμασίας περί τινα μέρη γινομένης, ἐπεγειρομένοις δὲ ταῦτα φανερὰ τοῦτον ἔχοντα τὸν τρόπον: ὥστ᾽ ἐπεὶ μικραὶ πάντων αἱ ἀρχαί, δῆλον ὅτι καὶ τῶν νόσων καὶ τῶν ἄλλων παθημάτων τῶν ἐν τοῖς σώμασι μελλόντων

γίνίεσθαι.

φανερὸν

καταφανῆ

μᾶλλον

*

AA

x

^

μὴν

οὖν ὅτι ταῦτα ἢ ἐν ^

καὶ

M

Evid

τῷ FA

ἀναγκαῖον

a

3

x

APXAS

κὺ

civai

εἶναι

^

ye

τῶν

καθ᾽

E]

{74

ὕπνον

^

τῶν

μεθ᾽

E

-

?

FA

αἴτια



πράξεων

διὰ

διάνοιαν ἐν οὖν ἐνδέχεται

x

x

τοῖς τῶν

Beare understands καταρρέοντος (463al5) to mean ‘running down the oesophagus’. Similarly Hett, ‘slipping down their throats’. Some mss read ἀπορρέοντος, ‘running off, the nuance of which will suit the supposed illusion of taste, if that is taken to require contact with the tongue. Reading οἷον at 463b2, following Beare and Hett.

ἀλλὰ, read by Ross, has stronger

ms authority, but is much harder to construe. The sea-battle and other far-off events do not contrast with a class of events where no causal initiative lies within the dreamers, but are illustrations of that class.

25

τὸ

τὰ δὲ πολλὰ συμπτώμασιν ἔοικε, μάλιστα δὲ τά τε ὑπερβατὰ πάντα καὶ ὧν μὴ ἐν αὑτοῖς ἡ ἀρχή, οἷον περὶ ναυμαχίας καὶ τῶν πόρρω συμβαινόντων ἐστίν: περὶ γὰρ τούτων τὸν αὐτὸν τρόπον ἔχειν εἰκὸς ὃν ὅταν μεμνημένῳ τινὶ περί

23.

20

La

φαντασμάτων

ἄλογον: ὥσπερ γὰρ ὄντες ἢ πεπραχότες καὶ πράττομεν (αἴτιον ἀπὸ τῶν μεθ᾽ ἡμέραν καθ᾽ ὕπνον κινήσεις

ἡμέραν

προωδοποιῆσθαι πάλιν καὶ τούτων τὴν davitrdopact τοῖς νυκτερινοῖς. οὕτω μὲν ἐνυπνίων ἔνια καὶ σημεῖα καὶ αἴτια εἶναι.

22.

15

ἐγρηγορέναι.

εἶναι τῶν οἰκείων ἑκάστῳ πράξεων οὐκ μέλλοντες πράττειν ἢ ἐν ταῖς πράξεσιν πολλάκις εὐθυονειρίᾳ ταύταις σύνεσμεν δ᾽ ὅτι προωδοποιημένη τυγχάνει ἡ κίνησις ἀρχῶν), οὕτω πάλιν ἀναγκαῖον καὶ τὰς πολλάκις

ἐν τοῖς ὕπνοις

10

30

463b

4638

10

15

20

ON DIVINATION THROUGH

SLEEP

109

experts say that one should pay extremely close attention to dreams. And that is a reasonable supposition even for those who are not practitioners, but are pursuing a theoretical inquiry. For movements occurring in the daytime, unless they are very big and powerful, pass unnoticed alongside those of the waking | state, which are bigger. But during sleep the opposite happens. For then even slight movements seem to be big. This is clear from frequent occurrences in the course of sleep. People think it is lightning and thundering, when faint echoes are sounding in their ears; or that they are enjoying honey and sweet flavours, when a tiny drop of phlegm is | running down;” or that they are walking through fire and feeling extremely hot, when a slight warmth is affecting certain parts. But as they wake up, it is obvious to them that those things have the above character. Thus, seeing that the beginnings of all things are small, so too, clearly, are those of illnesses and other affections imminent in our bodies. | Plainly, then, these must be more evident during periods of sleep than in the waking state. 47. 463a21-31. Dreams may also be causes. For just as our actions are often rehearsed in dreams, so conversely dreams may prefigure our subsequent acts. Thus, some dreams may be both signs and causes.

25

30

463a21. Furthermore, in the case of at least some appearances (phantasmata) in the course of sleep, it is not unreasonable to hold that they are causes of actions corresponding to each. For just as, when we are going to act, or are engaged in acting, or when we have acted, we are frequently | busy with these actions or perform them in direct dream-vision (the reason being that the movement has had the way paved for it by day-time origins), so conversely, movements in the course of sleep must frequently be origins of day-time actions, because our intention to do these has also had the way paved for it l'in the appearances (phantasmata) at night. In these ways, then, it is possible for some of the dreams to be both signs and causes. 48. 463a31-b11. But most fulfilled dreams may be put down to coincidence. Just as mention of a person need be neither a cause nor a sign of his turning up, so a dream may be mere coincidence in relation to its fulfilment. Hence dreams are frequently not fulfilled.

463b 463a31. Most of them, however, would seem to be all that are outlandish or concern matters where no dreamers themselves, e.g.?? a naval battle or things with these, matters probably stand as they do

coincidences, | causal initiative taking place far when a person

especially lies in the away. For mentions

110

ΠΕΡῚ

THE

KAO’

YHNON

MANTIKHE

463b

τινος TUX τοῦτο | γιγνόμενον: τί γὰρ κωλύει Kal ἐν τοῖς ὕπνοις οὕτως; μᾶλλον δ᾽ εἰκὸς πολλὰ τοιαῦτα συμβαίνειν. ὥσπερ οὖν οὐδὲ τὸ μνησθῆναι περὶ τοῦδε σημεῖον οὐδὲ αἴτιον τοῦ παραγενέσθαι αὐτόν, οὕτως οὐδ᾽ ἐκεῖ τοῦ ἀποβῆναι τὸ ἐνύπνιον τῷ ἰδόντι οὔτε σημεῖον οὔτ᾽ αἴτιον, ἀλλὰ σύμπτωμα. διὸ καὶ πολλὰ τῶν ἐνυπνίων | οὐκ ἀποβαίνει: τὰ γὰρ συμπτώματα οὔτε ἀεὶ οὔθ᾽ ὡς ἐπὶ τὸ πολὺ γίγνεται. 2. Ὅλως δὲ ἐπεὶ καὶ τῶν θεόπεμπτα μὲν οὐκ ἂν εἴη τὰ ΄

,

χάριν.

σημεῖον 3

δέ:

΄

ig

^

γὰρ 3

ὡς

οὐ

x

καὶ

bi

φύσις

ἐστὶ

t

μέντοι:

πάνυ t

εὐθυόνειροι, ἢ

,

δαιμόνια

^

,

γὰρ

3

a

ἀλλ᾽

Η

δαιμονία,

ἄνθρωποι

,

πέμποντος,

μελαγχολική,

,

φύσις

εὐτελεῖς ^

θεοῦ

10

ἄλλων ζῴων ὀνειρώττει τινά, ἐνύπνια, οὐδὲ γέγονε τούτου

x



5

+

ἀλλ᾽

Ν



προορατικοί

Y

[72

ὅσων x

3

àv

wv

παντοδαπᾶς

t

ὄψεις

,

|

Bela.

εἶσι

»

ὥσπερ

>

εἰ

15

καὶ 4

λάλος

^

x

ὁρῶσιν:

διὰ

γὰρ τὸ πολλὰ καὶ παντοδαπὰ κινεῖσθαι ἐπιτυγχάνουσιν ὁμοίοις θεωρήμασιν, ἐπιτυχεῖς ὄντες ἐν τούτοις ὥσπερ ἔνιοι ἀρτιάζοντες: ὥσπερ γὰρ καὶ λέγεται "àv πολλὰ βάλλῃς, ἄλλοτ᾽ ἀλλοῖον βαλεῖς", καὶ ἐπὶ τούτων τοῦτο συμβαΐνει. ὅτι

δ᾽

t^

ouk

4

ἀποβαίνει

4

3

TOAAa

,

τῶν

^

^

ἐνυπνίων, 5

οὐδὲν

LA

Ἷ

ἈΝ

20

ἄτοπον: Y

οὐδὲ γὰρ τῶν ἐν τοῖς σώμασι σημείων καὶ τῶν οὐρανίων, οἷον τὰ τῶν ὑδάτων καὶ τὰ τῶν πνευμάτων. ἂν γὰρ ἄλλη κυριωτέρα ταύτης συμβῇ κίνησις, Ab’ ἧς μελλούσης ἐγένετο τὸ σημεῖον, οὐ γίνεται: καὶ πολλὰ βουλευθέντα καλῶς τῶν πραχθῆναι δεόντων διελύθη δι᾽ ἄλλας κυριωτέρας ἀρχάς. ὅλως γὰρ οὐ πᾶν γίνεται TO

x

ὅμως Uu

24.

μελλῆσαν, ^

ἀρχάς 3

΄

οὐδὲ +

| γέ

x

Ζ

TO

^

τινας

AUTO +

x

TO

x

λεκτέον Ζ

ἐσόμενον 3

καὶ

,

εἶναι T

ad’ +

^

3

ὧν ?

TO

μέλλον’

x

GAA

+

οὐκ +

3

ἐπετελέσθη, ı

Ζ

Ross's text at 463b16-18 has been retained, though ὁρῶσιν in bl8 is grammatically difficult. There is some ms support for reading ὁρώντων, continuing the genitive participle construction governed by ὡς in b16. Whatever the correct reading, the sense required is as given in the translation. Beare and Hett both take the fact that ordinary people have God, but are seen by hardly so strong as precognitive dreams

prevision in dreams to ‘imply that such dreams are not sent by people of a certain nature’. But the connection marked by ὡς is an ‘implication’. It suggests, rather, that the occurrence of to ordinary people is what one would expect if they were due

not to God but to the dreamer's own nature.

25

+

30

463b

10

ON DIVINATION THROUGH

SLEEP

111

something, and then finds this very thing | coming to pass. What is to prevent this happening during periods of sleep as well? Rather, such things probably do happen often. So then, just as mentioning a particular person is neither a sign nor a cause of his turning up, so in the matter at hand, the dream is neither a cause nor a sign of its fulfilment for the person who experienced it, but is coincidence. That is also why many dreams | are not fulfilled. For coincidences happen neither always nor for the most part.

Chapter Two 49. 463b12-22.

Dreams are not sent by God. They are, however,

‘daemonic',

just as nature is. This is evidenced by people whose precognitive powers can be attributed to their nature. Since they have many visions of all kinds, they can be expected to strike lucky now and again. 463b12. In general, since some of the other animals dream, dreams could not 15

20

be sent by God, nor do they occur for that purpose. Nevertheless, they are daemonic. For nature is daemonic, but not | divine. Here is proof: quite ordinary people have powers of prevision and direct dream-vision, as if it were not God who sends dreams, but as if those whose nature is garrulous, as it were, or atrabilious see visions of all kinds.?^ For it is because they experience many movements of every kind that they just happen to encounter sights resembling real events, being fortunate in those, | like certain people who play at odds and evens. For, as the saying goes, ‘if you make many throws, you will throw differently on different occasions'. That is what happens in their case also. 50. 463b22-31. The non-fulfilment of dreams is not surprising. The events they prefigure may fail to occur, because more powerful causes intervene. For what is going to happen and what actually will happen are not the same.

25

30

463b22. That many dreams are not fulfilled is in no way surprising. The same holds for many signs of bodily events or of the weather, e.g. of rain or wind. | For if another movement should take place, prevailing over the one from which (when it was going to happen) the sign occurred, then the latter movement does not occur. And many well-made plans for actions needing to be carried out have been undone because of other causes that prevailed. For, in general, not everything that was going to happen actually does happen; nor is what will be the same as what is going to be But even so, | one should say

112

καὶ

ΠΕΡῚ

σημεῖα

πέφυκε

ΤῊΣ

ταῦτά

ΚΑΘ᾽

τινων

YIINON

οὐ

4630

MANTIKHZX

γενομένων.

περὶ δὲ τῶν μὴ τοιαύτας ἐχόντων ἀρχὰς ἐνυπνίων οἵας εἴπομεν, ἀλλ᾽ ὑπερορίας ἢ τοῖς χρόνοις ἢ τοῖς τόποις ἢ τοῖς μεγέθεσιν, ἢ τούτων μὲν μηδέν, μὴ μέντοι γε ἐν αὑτοῖς ἐχόντων τὰς ἄρχας τῶν ἰδόντων τὸ ἐνύπνιον, εἰ μὴ γίνεται τὸ προορᾶν ἀπὸ συμπτώματος, τοιόνδ᾽ ἂν εἴη μᾶλλον ἢ ὥσπερ λέγει Δημόκριτος εἴδωλα καὶ ἀπορροίας αἰτιώμενος. ὥσπερ γὰρ ὅταν κινήσῃ τι τὸ ὕδωρ ἢ τὸν ἀέρα, τοῦθ᾽ ἕτερον ἐκίνησε, καὶ παυσαμένου ἐκείνου συμβαίνει τὴν τοιαύτην κίνησιν προϊέναι μέχρι τινός, τοῦ κινήσαντος οὐ πάροντος, οὕτως οὐδὲν κωλύει κίνησίν twa | καὶ αἴσθησιν ἀφικνεῖσθαι πρὸς τὰς ψυχὰς τὰς ἐνυπνιαζούσας (ἀφ᾽ ὧν ἐκεῖνος τὰ εἴδωλα ποιεῖ καὶ τὰς ἀπορροίας), καὶ ὅποι δὴ ἔτυχεν ἀφικνουμένας μᾶλλον αἰσθητὰς εἶναι νύκτωρ διὰ τὸ μεθ᾽ ἡμέραν φερομένας διαλύεσθαι μᾶλλον (ἀταραχωδέστερος γὰρ ὁ ἀὴρ τῆς νυκτὸς διὰ τὸ νηνεμωϊτέρας εἶναι τὰς νύκτας), καὶ ἐν τῷ σώματι ποιεῖν αἴσθησιν διὰ τὸν ὕπνον, διὰ τὸ καὶ τῶν μικρῶν κινήσεων τῶν ἐντὸς αἰσθάνεσθαι καθεύδοντας μᾶλλον ἢ ἐγρηγορότας. αὗται δ᾽ αἱ κινήσεις φαντάσματα ποιοῦσιν, ἐξ ὧν προορῶσι τὰ μέλλοντα καὶ περὶ τῶν τοιούτων. οὐ

καὶ τοῖς

σοφοῖς,

διὰ ταῦτα συμβαίνει τὸ παΐθος τοῦτο τοῖς φρονιμωτάτοις. μεθ᾽ ἡμέραν τε γὰρ ἐγίνετ᾽ εἰ

θεὸς

ἦν



πέμπων:

οὕτω

δ᾽

εἰκὸς

τυχοῦσι καὶ ἂν καὶ τοῖς

τοὺς

4648

10

15

20

τυχόντας

προορᾶν: ἡ γὰρ διάνοια τῶν τοιούτων οὐ φροντιστική, ἀλλ᾽ ὥσπερ ἔρημος καὶ κενὴ πάντων, καὶ κινηθεῖσα κατὰ τὸ κινοῦν ἄγεται. τοῦ δ᾽ ἐνίους | τῶν ἐκστατικῶν προορᾶν αἴτιον ὅτι αἱ

25

4630

ON DIVINATION THROUGH SLEEP

113

that there are causes of a certain kind, from which no fulfilment ensued, and

these things are natural signs of certain things that failed to occur. 51. 463b31-464a19. Precognitive dreams above way may be explained as follows. water or air may be propelled towards would travel more easily at night and during sleep. 464a 463b31.

that cannot be accounted for in the Impulses akin to the movements of sleeping persons. Such movements would be perceived more readily

As for dreams that do not | contain causes of the kind we have

described, but outlandish ones, either in time or place or magnitude,

5

10

15

or in

none of those ways, but where those experiencing the dreams still have no causal initiatives within themselves, the following account (unless the prevision is due to coincidence) | would seem preferable to that of Democritus, who attributes it to images and emanations. When something has moved a portion of water or air, and this in turn has moved another, then even when the initial impulse has ceased, it results in a similar sort of movement continuing up to a certain point, although the original mover is not present. In this way it is possible that some sort of movement | and perception reaches the souls of dreamers, coming from the objects whence Democritus derives his images and emanations. And wherever they arrive, they may be more perceptible at night, because those carried by day are more easily dissipated (the air being less disturbed at night, because of | the nights being calmer); and thus they produce perception in the body because of sleep, people asleep being more sensitive to even slight internal movements than those awake. It would be these movements that cause appearances (phantasmata) from which people have previsions of what is going to happen, even on matters of the sort mentioned above. 52. 464a19-27. Such a theory would also explain why precognition occurs to ordinary rather than clever people. Their minds, being passive and vacant, are more responsive to external stimuli, as are those of the insane.

20

464219. This would also explain why this affection happens | to random subjects and not to the most intelligent. For if the sender were God, it would happen in the day-time and to clever people. But on our account, one would expect it to be random subjects who have prevision. For the mind of such a person is not reflective, but is deserted, as it were, and completely vacant.

25

| Thus, once set in motion, it is led on according to the direction of its moving

impulse.

The reason why some insane people have prevision is that their

114 3

TIEPI ^

οἰκεῖαι ξενικῶν

THX

KAO’

YIINON ^

κινήσεις οὐκ ἐνοχλοῦσιν οὖν μάλιστα αἰσθάνονται.

ἀλλ᾽

ΜΑΝΤΙΚΗΣ 1

464a

ἀπορραπίζονται:

^

τῶν

τὸ δέ τινας εὐθυονείρους εἶναι καὶ τὸ τοὺς γνωρίμους περὶ τῶν γνωρίμων μάλιστα προορᾶν συμβαίνει διὰ τὸ μάλιστα τοὺς γνωρίμους ὑπὲρ ἀλλήλων φροντίζειν: | ὥσπερ γὰρ πόρρω ὄντων τάχιστα γνωρίζουσι καὶ αἰσθάνονται, οὕτω καὶ τῶν κινήσεων: αἱ γὰρ τῶν γνωρίμων γνωριμώτεραι. οἱ δὲ μελαγχολικοὶ διὰ τὸ σφοδρόν, ὥσπερ βάλλοντες πόρρωθεν, εὔστοχοί εἰσιν, καὶ διὰ τὸ μεταϊβλητικὸν ταχὺ τὸ ἐχόμενον φαντάζεται αὐτοῖς: ὥσπερ γὰρ τὰ Φιλαινίδος ποιήματα καὶ οἱ ἐμμανεῖς ἐχόμενα τοῦ ὁμοίου λέγουσι καὶ διανοοῦνται, oiov ᾿Αφροδίτην φροδίτην, καὶ οὕτω συνείρουσιν εἰς τὸ πρόσω. ἔτι δὲ διὰ τὴν σφοδρότητα | oùk ἐκκρούεται αὐτῶν ἡ κίνησις ὑφ᾽ ἑτέρας κινήσεως. τεχνικώτατος

δ᾽

ἐστὶ

κριτὴς

ἐνυπνίων

ὅστις

δύναται

Αἱ 464b2 most mss give Φιλαιγίδου, but EM have Φιλιππίδου. No poet is known of either name. Leonicus Thomaeus, editor of the Juntine edition of the PN (Florence

26.

464b

τὰς

ὁμοιότητας θεωρεῖν: τὰς yàp εὐθυονειρίας κρίνειν παντός ἐστιν. λέγω δὲ τὰς ὁμοιότητας, ὅτι παραπλήσια συμβαίνει τὰ φαντάσματα τοῖς ἐν τοῖς ὕδασιν εἰδώλοις, καθάπερ καὶ πρότερον εἴπομεν. ἐκεῖ δέ, ἂν πολλὴ γίγνηται ἡ κίνησις, οὐδὲν ὁμοία γίνεται ἡ ἔμφασις καὶ τὰ εἴδωλα τοῖς ἀληθινοῖς. δεινὸς δὴ τὰς ἐμφάσεις κρίνειν εἴη ἂν ὁ δυνάμενος ταχὺ διαισθάνεσθαι καὶ συνορᾶν τὰ διαπεφορημένα καὶ διεστραμμένα τῶν εἰδώλων, ὅτι ἐστὶν ἀνθρώπου ἢ ἵππου ἢ | ὁτουδήποτε, κἀκεῖ δὴ ὁμοίως τί δύναται τὸ ἐνύπνιον τοῦτο. ἡ γὰρ κίνησις ἐκκόπτει τὴν εὐθυονειρίαν.

25.

30

1527), plausibly conjectured Φιλαινίδος, which makes the text refer to an

independently attested poetess, Philaenis, and which is supported by the letter v in the reading of L (Φιλαιγνίδου). In b3 some mss give, while others omit, the meaningless φροδίτην after ᾿Αφροδίτην. Various emendations have been proposed, but none is necessary, given the convincing explanation of the passage by Ross (283-4), which has been followed in the note on 464a27 (pp. 169-70). The text and the translation follow Ross in reading τύ with one ms at 464b15. Alternatively, reading τι with ELM, and accepting Biehl's emendation τοιοῦτο for the mss' τοῦτο, we should translate "Likewise in the case before us, the dream has some such effect’: i.e. dreams can exhibit distortion somewhat similar to that of a broken reflection.

10

15

4648

ON DIVINATION THROUGH

SLEEP

115

own movements do not block the access of others but get beaten off. Hence

they are especially sensitive to alien movements. 53. 464a27—bS5.

The experience of direct dream-vision by certain individuals

can also be explained. For example, friends, owing to their mutual concern,

are specially receptive to stimuli coming from each other. of atrabilious people can also be understood.

The precognition

464a27. The fact that it is certain people who have direct dream-vision, and

that familiar friends have prevision especially about each other, is due to their being specially concerned on each other's behalf. | For just as they are quickest to recognize and perceive one another at a distance, so too with their movements. For the movements of those familiar to them are correspondingly more familiar. Again, atrabilious people, because of their intensity, are good guessers, shooting, as it were, from a distance. Because of 464b their mutability, | the succeeding fancy (phantazetai) comes quickly before them. For just as even madmen utter or mentally rehearse things associated by assonance, e.g. ‘Aphrodite’, '-phrodite', as in the poems of Philaenis,?5 so do these people string a series onwards. And again, because of their 5 intensity, | one movement does not get knocked out of them by another. 30

54. 464b5-16. The most skilled interpreter of dreams is one who can observe resemblances, like someone who can reconstruct the original of a reflection from fragmentary or distorted images in troubled water. 464b5. But the most skilled interpreter of dreams is one who can observe resemblances. For anyone can interpret direct dream-visions. By resemblances, I mean that the appearances (phantasmata) are akin to images 10

15

in water, as indeed | we have said before. In that medium, if there is much disturbance, the reflection becomes in no way similar, nor do the images

resemble real objects at all. Indeed, it would take a clever interpreter of reflections to be able to detect readily and to comprehend the scattered and distorted fragments of images as being those of a man, or a horse, | or whatever. Likewise in the case before us, of grasping what this dream signifies.2° For direct dream-vision is erased by the movement.

116

ΠΕΡῚ

ΤῊΣ

ΚΑΘ᾽

YIINON

ΜΑΝΤΙΚῊΣ

4640

τί μὲν οὖν ἐστιν ὕπνος καὶ τί ἐνύπνιον, καὶ διὰ τίν᾽ αἰτίαν " 4 , ἑκάτερον αὐτῶν4 γίνεται, ἔτι δὲ περὶ τῆς ἐκ τῶν ἐνυπνίων μαντείας εἴρηται περὶ πάσης. ,ὔ

27.

*

#

Ross's text at 464018 has been printed, following Beare.

Some mss, followed by

Hett, omit περὶ πάσης. With that text, the line simply means ‘and again divination from dreams has been discussed'. The claim that so brief an essay has discussed 'every form' or 'every aspect' of divination from dreams seems extravagant. But the sentence may be an editorial addition rather than Aristotle's own. 28.

Most mss contain a few further words after 464bi8, introducing a work on the movement

of animals:

περὶ

δὲ

κινήσεως

Tfjg

κοινῆς

τῶν

ζώων

λεκτέον, ‘we

must now discuss movement that is common to animals. As Nussbaum (9) has noted, they give grounds for associating the De Motu Animalium with the Parva Naturalia.

464b

ON DIVINATION THROUGH SLEEP

117

55. 464b16-18. Retrospect.

464b16. It has been stated, then, what sleep and dreams are, and from what cause each of them occurs; and again, divination from dreams, in every form

of it," has been discussed.28

118

453b11

Notes ON SLEEP Chapter One 1. 453b11-24 The questions raised at b11-17, b17-20, and b20-24 will be taken up in the De Somno, De Insomniis, and De Divinatione, respectively. The De Somno The program outlined at b11—17 is followed closely in the remainder of the DS. For proof of this, see Lowe (esp. 285). He has shown that one passage, 455b13-34, is an insertion from an alternative, and probably earlier, draft of the

work, and that the concluding resumé (458a25—32) is also extraneous to the main treatise. With these two passages excluded, the overall structure is as follows. Aristotle shows (1) what 'affection' sleeping and waking are, by explaining (2) what part of the animal is affected in these states, and (3) from what cause each of them arises. The specific questions raised in b11—17 will be answered as follows: (8) sleeping and waking are common to body and soul (453b24—454a11);

(b) the part of the soul affected is the 'perceptual' part (455a4—b13), and the relevant bodily organ is the heart (455b34-456a29); (c) the cause of sleep is given at 456a30—458a10; (d) the cause of waking is given at 458a10-25; (e) both sleep and waking must belong to all animals (454a11—45523). (a) and (b) fulfil (2), while (c) and (d) fulfil (3), in the program outlined above. (1) is fulfilled by specifying what kind of affection sleep and waking are (viz. states of perceptual incapacitation and functioning), and by describing the particular cause that distinguishes sleep from other kinds of sensory incapacitation. On this view, the attempt to explain sleep in terms of a 'final cause' (cf. Glossary, s.v. aitia) belongs to extraneous material that formed part of an earlier draft. See note 8 on 455b13—34. The De Insomniis Aristotle will answer the questions of b17-20 as follows: (a) dreams are the work of the imagination (to phantastikon) (ch. 1). They are presentations due to traces of perception that reach awareness during and because of sleep (chs. 2-3).

453b18

ON SLEEP

119

(b) The occurrence or non-occurrence of dreams will depend upon whether the sleeper's internal state is or is not favourable to their appearance (461a8—b7). (c) It is not the case that people always dream, yet fail to remember doing so. Sometimes their physical state will prevent dreams from occurring at

all (cf. 46131721). b18-19. The remember' remember Malcolm when

hypothesis that 'people always anticipates the view, prevailing in by far the greater part of what we (see Introduction, section 10), the

asleep,

but

do

not

remember

dream when asleep, but do not modern psychology, that we fail to dream. By contrast, on the view of hypothesis that we always dream

having

done

so,

must

be

incoherent.

Aristotle does not view the hypothesis as incoherent, nor does he endorse it himself. Rather, he advances a theory of dreams according to which it 15 empirically false. For he holds that under certain conditions explained at 461a17-21, we do Introduction, 50-1.

not

dream

at all.

See

note

43

on

462a31-b11

with

The De Divinatione In answer to the questions of b20-24, (a) Aristotle will allow that foreknowledge may be gained from dreams, but only from those that are 'signs' or 'causes'. Otherwise, veridical dreams are put down to coincidence (462b26—

463b11). (b) Foreknowledge from dreams need not be limited to events brought about

by human agency (cf. 463a4-21, 463b12-31, 463b31—-464b5). b22-23.

Why

does Aristotle ask whether prevision is limited to future human

actions or covers a wider range of events? Presumably, a restriction of it to the

former would be congenial to those who ascribed dreams to the subject's waking preoccupations, and would therefore have conceded that dreams might give foreknowledge of the dreamer's own actions. Aristotle will recognize dreams of this type himself (463a21—31), and will allow that dreams may bring about their own fulfilment, by suggesting a course of action to the dreamer. In addition to these, he will allow dreams to be 'signs' of things not within human control, such as the dreamer's own state of health (463a4—21). Other veridical dreams

he ascribes to coincidence.

They

do not, for him,

evidence

genuine

precognition, but may be due to the dreamer's peculiar mentality (463b15-19,

464a17-b5). b23-24. Note Aristotle's expression for the subclass of future events other than those due to human agency. The words translated 'things for which the daemonic is responsible’ mean, more literally, ‘things of which the daemonic has the cause (aitia). They are explicated by 'things that happen by nature or fortuitously’.

120

NOTES

453b24

Thus, natural or fortuitous events specify the range of things attributable to the daemonic. On this interpretation, 'the daemonic' covers, and in some sense explains, all events not due to human agency. Beare's translation, 'supra-human agency’, suggests that Aristotle is postulating a supra-human agent responsible for natural or fortuitous events. But this need not be the implication of the word 'daemonic', either here or at 463b14, where it is predicated of both dreams and nature. See Glossary, s.v. daimonion, note 49 on 463b12-22, and Introduction,

section 10. In ‘things that happen by nature or fortuitously’ (024), the fortuitous is contrasted, not equated, with what happens by nature. But fortuitous events occur, for Aristotle, within the natural order. For on his account of ‘the fortuitous’ in Phys. IL. 6, events due to it are themselves part of nature. The

point of attributing an event to the fortuitous, is not to appeal to ‘the supernatural’, but to deny that it occurred by design, even though it gave the appearance of having done so. See Glossary, s.v. automaton.

2. 453b24-454a11 Aristotle here answers the first two (1) that sleeping and waking ‘belong of the animal's perceptual capacity, viz. (as will appear later) the heart; psychophysical states, proprietary to

questions posed at 453b12—14, by arguing to the same part’, i.e. they are both states and of the master-organ that controls this, (2) that sleeping and waking are both alike neither soul nor body alone.

b24—31. The argument turns upon the conception of sleep and waking as 'opposites', the former being a 'privation' of the latter. The opposition of sleep and waking is traditional.

Cf. Heraclitus (DK 22) frs.

B26, B88, and Plato, Phaedo 71c-d.

With respect to sleeping and waking, Aristotle will remark later (462a26-27) that one of these states may be present in an unqualified sense while the other is present ‘in a certain manner. This allows for a subject to be fully asleep, yet sufficiently awake to have some perceptual awareness. In general, however, Aristotle defines sleep in terms of the absence of actual perception (454b9-1 1), and waking in terms of the power of perception being ‘set free’ (454a32—b1).

Thus there is a conceptual connection between

sleep and the absence of

perception. See also notes 21 on 458b3—9, 42 on 462a15-31. For the concept of 'privation', see Glossary s.v. steresis. ΤῸ view sleep as 'a sort of privation of waking' (b26-27) is to regard it as an absence (in this case only temporary) of a positive condition. Other examples of 'privation' concepts are given at b29—31,

e.g. blindness and deafness.

These

are to be understood

likewise as lacks of their positive counterparts. b27-31. The principle invoked is that where F and G are opposites, then both F and G must inhere in a common subject. Hence, if sleep and waking are opposites, then what they are states of must be one and the same subject.

454a8

ON SLEEP

121

Moreover, that subject cannot, in the case of sleep and waking, be either the soul or the body alone. It is not possible, then, for the soul to be awake, while the body is asleep, or conversely. In effect, Aristotle is here rejecting a dualistic view, according to which sleep is a condition of the body, from which the soul -itself remaining awake and active — may temporarily withdraw. Aristotle, following Plato, had once held a dualistic view himself. See Introduction, 14. Against it he now argues that it must be the same subject that is said to be asleep or awake, hence the ascription of those states to two different subjects, body and soul respectively, must be incorrect. Neither state can be intelligibly be ascribed to soul or body alone. It must therefore be wrong to suppose that in dreaming the soul has become 'separated' from the sleeping body and has vision in a disembodied state. It would be equally incorrect to suppose that sleep could characterize a body bereft of soul, viz. an inanimate object. For such a body would be incapable of perceiving (b11), and therefore of being asleep. Since we cannot ascribe the exercise of perception to inanimate things, we cannot ascribe perceptual powers to them either. But things lacking such powers can no more be asleep than they can be awake. So the idea of a sleeping body from which the soul had withdrawn would be absurd. a8.

What owns any capacity also owns its exercise: This principle is needed to argue that if the exercise of a capacity (viz. actual perceiving by a waking subject) requires both body and soul, then so does the mere possession of that capacity. Hence, if sleep entails the possession of the capacity for perceiving, then sleep no less than waking requires both body and soul, i.e. it can belong only to an animate substance.

3. 454ali—b9 all-12. We have already distinguished elsewhere: i.e. in DA II. 2-3. In arguing that plants can neither sleep nor wake because they lack 'the perceptual part', Aristotle sets aside the question whether that part is separable from others. At a14 he says that none of the other parts can exist separately from the nutrient, whereas the question whether the perceptual part is separable is waived, with the concession that it actually is separable 'as a capacity and in its being' (418-19). By this concession Aristotle means that even though perception is nowhere found, and cannot exist, without the nutrient part, nevertheless the former capacity is a different one, i.e. it is distinct in what it is. A similar point will be made later with respect to the powers of perception and imagination, which are said to be 'the same', yet to differ in their 'being' (459a15-17). Aristotle sometimes expresses such points by saying that two things may be inseparable (1.6. neither can exist independently of the other), and yet be separate ‘in account' (logo(i)). One and the same item may satisfy both a definition of X and a definition of Y, so that to talk of it as X and to talk

122

NOTES

454a19

of it as Y will be to speak of the same thing under two different descriptions. For

other

examples,

see DA

413b29-32,

425b26-27,

427a2-3,

432a31-b4,

433a21—25, Sens. 449b14-20, EN 1102a30-31, 1130a12, 1141b23-24. al9-b9.

Aristotle gives a proof that sleep and waking must characterize all animals.

With the emended text at a21 on which the translation is based (see textual note

1), the gist of a21—26 is as follows: (a) without perception no animal could either sleep or wake; (b) with perception every animal must do both. The point of (a) may be to glance at the condition of embryos, which is not sleep proper but more akin to the state of plants (cf.: GA 778b19—779a11). a26-b9. It is shown (1) that no animal could be always awake. It will be shown at 454b9-14 (2) that no animal could be always asleep. The argument for (1) invokes the notion of a 'function', for which see Glossary, s.v. ergon. For the idea that no function can be exercised continuously, cf. EN 1175a5, 1176b34. The thesis that sleep provides a necessary respite in the animal's performance of its functions appears to anticipate the later account at 455b16—28 (cf. 458a31-32) of the ‘final cause' of sleep. But if, as Lowe has argued, that later account is an insertion into the text of a fragment

from

an earlier draft of the De

reformulation

of the

idea

that

Aristotle's present purpose, i.e.

Somno, sleep

has

the present argument is merely a ‘final

cause’,

in order

a

to serve

to show that sleep is universal in animals.

See

Lowe, 284-7 with n. 33.

The argument runs as follows: (i) Waking and sleeping are opposites, such that an animal must possess one of them at any given time (b1—3, 453b24--26).

(ii) Being awake consists in, or is defined as, perceiving (a31-bl, cf. 454a1—5). (iii) Perceiving is the function of a certain part of the animal (a29—30). (iv) All parts that have a function are incapable of exercising it without respite (a26-29, cf. 455b18—19). From (ii), (ii), and (iv), it follows that, (v) Nothing could be awake at all times. From (1) and (v) it follows that, (vi) sleep at certain times is necessary for every animal. b4-9.

Aristotle argues that if excess of waking can occur without illness, so can the corresponding incapacitation or dissolution of the perceptual powers, in which sleep consists. Sleep, then, is in no way an unhealthy or unnatural state, but whatever is awake must necessarily be subject to it. In this respect it differs from other states of unconsciousness from which Aristotle will distinguish it (455b2-13, 456b9-11), and from the other cases of 'privation' mentioned at

453b29-31.

454b9

ON SLEEP

123

The foregoing argument illustrates how understanding of natural phenomena can be deepened, for Aristotle, by a grasp of the relevant concepts. Once it has become clear what sleeping and waking essentially are, the need for every animal to sleep can be recognized a priori. Thus for sleep, as for dreams (see note 42 on 462a15-31), common observation is reinforced by conceptual and scientific understanding. 4. 454b9-455a3 lt is now shown that nothing could be always asleep. This follows from Aristotle's account of sleep as the 'fettering' of the perceptual part. For whatever has the power of perception must be capable of exercising it at some time, yet it cannot be exercising it actively (i.e. 'exercising perception in the chief and unqualified sense', b13), while it is still asleep. Note that only full-blown perceiving is held to be precluded by sleep: cf. 455a9-10. The words ‘exercising perception in the chief and unqualified sense' are probably intended to leave room for dream awareness, which will be called ‘in a certain way, a sense-impression' (456226). It remains a problem, however,

to understand how this awareness is possible during sleep, while ordinary perception is not. See Introduction, sections 5, 6 and 11.

b14.

The conclusion that ‘all sleep must be subject to awakening’ should not, of course, be taken to mean that every individual sleeper must at some time wake up. Aristotle is not precluding the occurrence of death during sleep. Nor is he ruling out irreversible coma. At EN 1095b32~33 he remarks that someone might possess virtue, and yet remain asleep throughout his life. But such an individual's condition would differ from the ordinary kind of sleep under consideration here. Aristotle is not considering individual animals. He means only that there could be no species whose members spend all their lives asleep. For normal sleep is, by its nature, a temporary immobilization of the perceptual powers, enabling those powers to be exercised once more in waking life (cf. 455b16-28).

b14-23. Aristotle claims empirical confirmation for the thesis that sleep is universal. For the species mentioned, cf. HA 523b1-5 (molluscs), HA 525b37-526a9 (hard-eyed creatures, e.g. cray-fish), and HA 523b8-12 (testaceans). See, also, the empirical study of sleep in HA IV.10, printed in the Appendix. b22-23. But anyone who has found the foregoing argument convincing will be persuaded of this: With the emended text adopted (see textual note 4), 'this' refers to the view that testaceans sleep. Aristotle here invokes the a priori reasoning of 454a26—b9 to cover the one species (testaceans) for which he admits that empirical evidence is still lacking. b23-a3. The main conclusions of ch. 1 are resumed, with additional argument to show that sleep and waking cannot occur in plants. This is because they possess

124

NOTES

455a4

only the ‘nutrient’ part, whose independence of perception is indicated by its enhanced functioning during sleep. Cf. EN 1102b3-4.

Chapter Two Aristotle now argues that sleep must be an incapacitation of the primary sensefaculty, and the primary sense-organ, viz. the heart.

5. 455a4-12 Sleep cannot consist simply in failure of each of the special senses (sight, hearing etc.) independently of each other. For during sleep all of these are affected. It is not possible, while asleep, to perceive with one sense, but not with another, as could be expected, if sleep consisted in several independent sensory failures. a8.

The reference to the discussion of 'imperfect' animals is to DA 425a10, 432b20— 23, 433b31.

239-10. An animal, while sleeping, cannot be experiencing any perception at all in an unqualified way: Aristotle has already said (454b13-14) that full-blown perceiving during sleep is impossible. That is, for him, not merely a contingent fact, but follows from a correct understanding of what sleep is. As a 'privation' of waking perceiving

(453b26—27), which has itself been defined in terms of actual (454a4—5, 454a32-b1), sleep is, essentially, a certain sort of

temporary incapacity to perceive. The point of the reservation 'in an unqualified way' is to leave room for (a) dreams, which will later be treated as quasi sense— impressions (456426, cf. DI chs.2-3), and perhaps also (b) occasional faint sensory awareness attributed to sleepers at 462a19-27. 6. 455a12-b2 This passage is a key text for Aristotle's important but difficult concept of ‘common

sense’, so called at DA 425a27, DM

450a10-13a, and PA 686a31--32.

Aristotle does not use the phrase 'common sense' here, but speaks, rather, of a ‘common faculty’ (416). It is disputed whether this is to be equated with 'common sense', or whether the latter performs merely the specific role of apprehending 'common sensibles', i.e. objects common to more than one sense. For these objects cf. 458b4—6, and see note 21 on 458b3-9. A masterly account of the operations of 'common sense' in relation to the special senses and their respective organs is given by Kahn. See also Modrak, ch. 4, especially 62—77. Sleep is here explained as a temporary disabling of the 'common' faculty. At al7-20 Aristotle refers to two of the faculty's roles: (1) it is that whereby one is

aware that one is seeing and hearing; and (2) it is that whereby one can judge

455a12

ON SLEEP

125

that sweet things differ from pale ones. Arguments relevant to these roles occur at DA 425b12-25 and 426b8—427a16 respectively. The faculty is 'common', evidently, in that it attends upon the exercise of all of the special senses, accounting for inter alia the subject's awareness of seeing, hearing etc. (416-17). Just how it does this, however, remains elusive.

Aristotle's subtle view of the relation between the ‘common’ faculty and the special senses 1s crystallized in the difficult words at a20—22: 'For there exists a single sense-faculty, and the master sense-organ is single, though its being differs for the perception of each kind of thing, e.g. of sound or colour. The idea is elaborated in the De Sensu, where Aristotle says that objects of perception may be ‘one and the same’ and yet ‘differ in their being’ (449a14-16). For example, honey or wine may be both pale and sweet, and thus one and the same object may satisfy two different descriptions. The same thing is both pale and sweet, yet different accounts are required of what it is for it to be pale and what it is for it to be sweet. Analogously, Aristotle suggests (449a16-20), there is a single sense-faculty, yet each of the modes in which it operates (visual, auditory etc.) is different. A single sensory apparatus is capable of receiving data from a wide variety of external stimuli via (in the case of the higher animals) five quite different kinds

of receptor. Hence, different accounts of what it is (the ‘being’ of this apparatus) will be required for each of these modes of operation. Once the notion of a single sensory apparatus operating in various modes is grasped, Aristotle's view of the relation between the ‘common sense’ and the special senses becomes intelligible, in spite of his mistaken identification of the ‘primary sense-organ' (i.e. the central organ of the apparatus) with the heart. As Kahn (21) observes, the view 'is easy enough for us to understand, if whenever Aristotle says "veins" or "channels" we substitute "nerves", and whenever he refers to the heart we think of the brain instead. Once these substitutions have been made there is nothing strange in his notion of a central organ serving as sensorium proper, the point at which all stimuli from the external organs converge and in which they must appear for any genuine sensation to occur’. Like a multipurpose tool, the brain might be regarded as a single thing, yet also as many things as the different activities which its various functions enables its owner to perform. A complete understanding of 'what the brain is' (its 'being) would require a differentiated account of its role in all these activities. But it remains a single, primary organ, ‘primary’ in that its functioning is essential for each and every part of our sensory apparatus (and much else within us) to work (cf. 455b11-13). That is Aristotle's view of the sensorium centred in the heart. A single power supply provides heat, light, and other services for a house. Yet separate accounts of heating, lighting, and other services, are needed, if we are to understand what the power supply is and how it functions. This analogy

126

NOTES 455a22 may serve to clarify the argument of a27-b2. Aristotle is arguing, in effect, for the existence of a single 'power supply', which is temporarily shut down during sleep. For only thus, he reasons, can we explain why, during sleep, all of the special 'services' should be knocked out at the same time.

a22-27.

The reference to an earlier discussion of touch is to DA 413b4-10, 415a3-6,

435a12-14.

The translation, ‘And this accrues to an animal at the same time as

the sense of touch, in particular', is most naturally taken to mean that as soon as you can ascribe touch to an animal, you can ascribe to it a master sense-organ. Because the common sensorium is specially associated, in some way left unexplained, with the sense of touch, the fact that all animals possess that sense, and

therefore

a common

waking and sleep. made by Woods

a29—30.

The

sensorium,

is proof that all animals

must

share

in

This revised translation and interpretation follow suggestions

(186-7).

senses

'need

not,

and

in

a certain

way

cannot,

be

exercised

simultaneously’. It is obvious that they ‘need not' be exercised simultaneously. Lying awake at night, one may feel the sheets, yet see, hear, taste, and smell nothing. It is less obvious that the senses 'cannot' be exercised simultaneously, as, for example, when eating in a noisy, smoke-filled room. Note, however, that

'cannot' is qualified by 'in a certain way'. Aristotle probably means concentrated perception by one sense tends to withdraw attention from objects of others (Ross, 259). The difficulty, in any case, does not affect Aristotle's main point, which is since the special senses can function independently of one another while we awake, this ought also to be possible during sleep, unless it were the case their operation depends upon a single master-organ.

that the that are that

7. 455b2-13 Sleep is now distinguished from other unconscious merely in sensory inactivity or in loss of capacity to also attend, e.g., fainting or throttling. Sleep unconsciousness that results from the disabling of the cause of that disabling will be given in chapter 3. b6.

states. It cannot consist perceive. For those states is the specific kind of primary sense-organ. The

Certain unconscious states: The meaning is uncertain, but the broad sense of ‘unconsciousness’ is required elsewhere (457b25), and suits the design of the treatise better than 'derangement' (Hett). See Glossary, s.v. eknoia and Lowe, 290, n. 17.

8. 455b13-34 This passage promises to consider the explanation of sleep in terms of Aristotle's celebrated doctrine of ‘four causes". Since the doctrine provides a general schema of fundamental importance for Aristotelian explanation, a brief

455013

ON SLEEP

word about it will be in order here.

127

For references and further remarks see

Glossary, s.v. aitia.

The doctrine may be roughly summarized as follows. Aristotle distinguishes four types of explanatory factor which may be relevant in seeking to account for some item, phenomenon, or state of affairs. For a complete scientific understanding,

we need to ask four different sorts of ‘why

...?'

question, and

(where appropriate) give four different sorts of 'because ...' answer. (1) We may explain the properties of an item in terms of its material composition (traditionally called its 'material cause). (2) We may show how its properties derive from its essential nature, or from a formal account of what it 1s (its ‘formal cause’, referred to in our text simply as ‘the account’). (3) We may attribute the item or its properties to the agent which originated, produced, or brought it about (its ‘efficient cause’, called in our text 'the source of change"). (4) We may invoke the purpose which the item serves, the goal at which it aims, or the function it performs (traditionally, its ‘final cause‘, here called the 'for the

sake of something’, or ‘the end"). The modern English ‘cause’ covers, at best roughly, only the ‘efficient cause’. When used for the other three, 'cause' is an archaism, a quaint relic of the Latin causa in which the doctrine was transmitted. The doctrine of 'four causes' will reappear in the closing lines of the De Somno (458a25-32). For the supposed ‘material cause' of sleep, see note 15 on 457a33-b6.

It is, however, doubtful whether the doctrine does, in fact, provide

a schema for understanding the treatise in its extant form, and whether Aristotle

continued to bold that sleep admits of all four types of explanation. Commentators have noticed that our passage is sharply discontinuous with what has preceded it. It was regarded by Drossaart Lulofs (1947, xv—xxvii) as the start of a separate version of the treatise on sleep, extending all the way to 458232, and completed earlier than the preceding material, which Lulofs took to be an unfinished revision. This hypothesis was rejected for good reason by Ross, 12, and Siwek, xiii-xiv. More recently, however, Lowe has shown that the break Lulofs discerned is genuine, but has rejected Lulofs' explanation for it. Instead, he argues

convincingly that the present passage (ending at b34) is a short fragment of an older version, in which Aristotle proposed to explain sleep in terms of his 'four cause' schema. This proposal he later had reason to abandon, in as much as he no longer believed in a 'final cause' for sleep. But an editor of his text preserved the fragment by merging it with the version of the De Somno that had superseded

it.

Likewise,

the closing lines of the treatise (458a25-32),

recapitulate its conclusions in terms of the 'four causes', are viewed an editorial adjustment, making the recapitulation suit the inserted this view is correct, the attempts of modern editors (e.g. Ross, 260) what follows the present passage fulfils its promise to consider all

which

by Lowe as passage. If to show that four 'causes'

128

NOTES of sleep are misguided. (455b16—28),

followed

455b15

All we are given is a discussion of the ‘final’ cause by two

discussions

of the ‘efficient’ cause,

begun

at

455b28 and 456a30 respectively. The ‘formal! and ‘material’ causes are nowhere to be found. For criticism of Lowe's view, see Woods, 180-1.

815.

The ‘for the sake of something’: ie. the ‘final cause’ (see Glossary, s.v. aitia). Such explanations are familiar for human actions performed in order to attain ends that their agents consciously pursue. But they are also appropriate, in Aristotle's view, for items and processes in nature that are not due to human

agency, and where no subject consciously pursues the ends they subserve. This ‘natural teleology’ pervades Aristotle's biological writings throughout, and is explicit in the statement (b17) that ‘nature acts for the sake of something', which is a premiss in the argument that sleep occurs for the sake of animal preservation (b22, cf. 458a31-32). See also Phys. II. 9, PA 639b19-640a10, 64241--13, GC 337b14-338a4.

A lucid account of Aristotelian natural teleology

is given by Owens. Teleological explanations discussed further in the Introduction, sections 7-8.

of sleep

and

dreams

are

b25-26. Hence it is necessary for sleep to belong to every animal: The translation deviates from Ross's text (see textual note 8) in taking this sentence as a corollary of what precedes it, and not as a further argument for the thesis that sleep preserves animals. 'Conditional necessity', according to which certain conditions are requisite if a certain goal is to be realized, applies to sleep precisely because it is needed for the animal to exercise its perceptual and intellectual powers in the waking state. It is not (as Ross's text would require) an independent ground for the thesis that sleep serves that purpose. For the doctrine of conditional (or 'hypothetical') necessity, see Cooper, 200-1, 209-16, esp. 210, n. 8, and Nussbaum, 88-93. For the thesis that activity without intermission is impossible, cf. the argument

of 454a26—32, and see EN 1154b7, 1175a4, 1176635, 1177222.

It is true that

some organic functions, such as respiration or digestion, continue without interruption. Aristotle himself notes that nutrient functions are exercised par excellence during sleep (454b32-455a3). The present thesis is affirmed, however, and is plausible, for perceptual and intellectual capacities whose exercise is attended by pleasure (b19). Modern studies of the effects of sleep deprivation have confirmed the thesis empirically. Perception and thought cannot continue indefinitely without severe impairment. b28-34. This passage announces, but does not carry through, an account of the 'efficient cause' of sleep, in keeping with the 'four cause' methodology enunciated at b13-16 above.

Note that the point made here, viz.

that in view of

the similarity of bloodless animals to sanguineous ones, and of the latter to human beings, ‘all cases must be studied in the light of these’ (033), would lead But the us to expect concentration upon human beings in what follows.

455034

ON SLEEP

129

succeeding passage (b34—456a29) is not restricted to human beings, nor does it, in fact, deal with the efficient cause of sleep at all. It merely argues that the heart is the seat of perception in all animals. This discontinuity marks b28-34 as part of the inserted fragment. See Lowe, 284—5.

9. 455b34-456a24 This passage picks up the exposition from 455b8-10 above, by showing which particular organ is responsible for sleep, thus differentiating it from other forms of perceptual incapacitation. The relevant organ must be the master-organ that controls both perception and movement, viz. the heart, or its analogue in bloodless creatures. a2-3.

412.

It has already been determined in other works: The reference is to PA 665a10-15, where it is argued that perception and movement must have a common source. Their

naturally

inherent breath:

For this term, see Ross, 40-3.

breath' belongs by nature to all animals that do not respire. them, i.e.

'Connate

It is internal to

not inhaled from outside.

a20-21. Given that it is upon the occurrence of some perception, either internal or external, in the primary sense-organ, that every animal moves: The meaning of these lines, and their exact role in the argument, are somewhat obscure. Modrak (137) cites them as showing that the physiological convergence of the special organs in a central organ enables Aristotle to explain the unity of perceptual experience. She translates: 'and in each case when a sensation, either internal or external, arises,

a movement

occurs in the primary

sense-organ'. Aristotle did, indeed, think of the special organs as converging upon a central one, and he did hold that impulses must reach the central organ for sensation to occur. But here his word-order suggests that it is the perception, rather than the movement, which occurs in the primary sense-organ; and his syntax suggests that the movement is dependent upon the perception, not vice versa. Movement’ in the surrounding context (456a2, a5-6, a18, a24) means an animal's locomotion or movement of some part of its body (a15). So here it probably means locomotion, rather than an impulse or stimulus to the primary sense-organ. This view is confirmed by the aside that follows (a24-26): the movements of sleepers (e.g. sleep-walking) are no exception, as might be supposed, to the principle that movement requires some sort of perception; for even they are ‘not without’ some sort of 'appearance' (phantasma) or perception (aisthesis), since a dream is a certain sort of sense-impression, The implication is that every movement of an animal, whether waking or sleeping, is a response to some internal or external stimulus. This interpretation is strengthened by reading genomenes at a20 with Ross, which allows us to understand the

130

NOTES movement as prompted by the perceptual stimulus, and concurrent with it. See textual note 9 and Beare's notes ad loc.

456a24 not as merely

10. 456a24—29 8.24. Some sleepers move: sc. in their sleep. 826.

For a dream is, in a certain way, a sense-impression: For 'sense-impression' see Glossary, s.v. aisthema.

As will appear in the De Insomniis, dreams are not

properly thought of as sense-impressions, but are akin to them. They are like regular sense-impressions, but are produced by a different kind of stimulus to the sense-organs. a27. The promise of a later discussion presumably looks forward to the De Insomniis, although Aristotle nowhere there considers why people remember their dreams but not the 'waking actions’ they performed in sleep (427-29). He is evidently thinking of actions in sleep akin to waking ones, e.g. sleep-walking, rather than ‘the actions of waking life’ (Ross, 259). His puzzle is represented by Huby (61, n. 9) as ‘the question why people remember dreams on waking but not vice versa’. Cf. also Beare (1906), 301 with n. 1. But this is a misunderstanding. At GA 779a16-19 Aristotle notes the high degree of awareness possessed by sleep-walkers. Since their actions are closely akin to waking behaviour (far more so than is much of their dream experience), it is indeed surprising that they should often be totally unaware of such episodes upon waking, yet will remember their dreams clearly. That, rather than the fact that we fail to remember our waking actions when dreaming, is what needs explaining. Aristotle here refers to an explanation ‘in the Problems’. The extant work of that name in the Aristotelian corpus contains no such explanation, but its authenticity has been widely and seriously doubted. Aristotle may therefore be referring to a lost work of the same name.

Chapter Three This chapter describes in detail the cause (aitia) of sleep, showing how it is produced by the movement of heat, following upon the ingestion of food. It gives an account of the ‘efficient cause’ of sleep, begun earlier (455b28), but not yet completed. See note 8 on 455b13-34. Much of the physiological detail remains obscure. Wiesner (269), after a thorough discussion, suspects that ‘Aristotle may have had difficulties in arriving at a definite conception of [the role of heat and cold] in connection with his doctrine of the heart as the principle of both perception and nutrition’.

456430 11. b2.

ON SLEEP

131

456a30-b28 The dissections: This may Alternatively,

refer to a work on anatomies no longer extant.

Aristotle means

that the connection

of veins

with the heart is

shown by actual dissections, carried out by himself or his students. See Ross, 264, Siwek, 201, n. 58. The part played by dissections, amongst other empirical methods, in the formation of Aristotle's physiological theories 1s discussed by G.E.R. Lloyd. For their bearing upon the theories of the DS, see Lloyd 216-7, 226-8. b6.

The work on nutrition referred to here is mentioned by Aristotle in several other places. It appears to have been lost, or was never completed. Cf. Siwek, 201, n. 59, Nussbaum 375-6.

b9-12.

For the distinction between

sleep and other kinds of unconsciousness,

cf.

455b2-13. b12—17. The difficulty raised here is elliptically expressed. For the translation 'to have fallen asleep' (b13), see textual note 10. "This point' (D12) must be the fact just mentioned, that appearances have been known to occur to people in a deep faint. What, Aristotle asks, could account for these? In reply, he suggests that ‘if it is possible for one who has fainted to have fallen asleep, it could also be that the appearance (phantasma) was a dream' (b13-15). But he means to reject that suggestion as incorrect, and the conditional should be read as counterfactual: if it were possible that the subject had fallen asleep, then his experience could be a dream; but that is not possible on the present account of sleep, which has differentiated

it from

fainting and other unconscious

states; hence,

such

experiences are not to be explained as dreams, and are still puzzling. Aristotle does not offer any solution to the puzzle. After contrasting the states in question with sleep proper (as at 455b2-13), he returns to his main theme at b17.

Note the assumption, here as elsewhere (459a19-20, 462a15-16, 462a30—

31), that being asleep is a necessary condition for dreaming, so that if a subject in a deep faint is not asleep, he cannot be having a dream. b15-16. Many things, too, are reported by those who have gone into a deep faint and have seemed dead: This translation assumes that the words are not uttered during the swoon, but upon regaining consciousness. Ross (264) compares Plato's story of Er (Republic 614b2-621b7), in which experiences occurring while the subject was supposedly 'dead' are described upon his return to life. For such phenomena Aristotle would like to be able to give a physiologically based explanation similar to his account of dreams, but he acknowledges here that they cannot be treated as dreams. Note that he says only that the subjects have 'seemed' dead, not that they had (like Er) actually died. b19.

The exhalation attendant upon nutrition: The literal meaning of the word translated ‘exhalation’ is ‘steaming up’, as from a swamp. ‘Evaporation’ (Ross,

132

NOTES

456b28

Hett) sounds less suitable for solid food, but the process of upward movement and downward reflux will be described later (457b30-458a5) by analogy with the evaporation of water and its descent in rainfall. Cf. also PA 652b34653a10. Note how Aristotle's theory, here as elsewhere, accommodates facts of common

observation,

e.g. heaviness

of the head, drowsiness

the development of catarrhs (458a3-4).

after meals,

and

For other examples, cf. 45967-23.

12. 456b28—457a3 b34-35. For fatigue is a solvent; and the dissolved matter, unless it is cold, acts like food not yet digested: The notion of 'dissolved matter' or 'colliquescence' (suntegma) reappears at GA 725a1—b3, where it is described as a useless and potentially harmful residue, which 'runs about in the body wherever it can find a clear way for itself". The solvent effect of fatigue is due to the heat that it generates. This acts upon residues in the stomach and intestines, and forces them upwards. The dissolved matter will then produce heaviness in the head, as do the soporific agents Aristotle has mentioned at b28—34. Cf. a2-3, where feverish and comatose

conditions

are attributed to moist

or hot residues,

and 462b29-30

where fatigue is cited as a cause of fever.

13. 457a3-21 a9.

In fact sleep is, in a way, an epileptic seizure: On epilepsy, cf. [Probl.] 953a5-19, 954b27-31, 960a18-19. ‘Epilepsy’ means, literally, a 'seizure', and the related word katalepsis will be used at 458229 of sleep. Aristotle thinks of sleep, like epilepsy, as a temporary condition, in which normal functioning is suspended. At EN 1150b32-35 a contrast between weakness of will and wickedness is illustrated by analogy with epilepsy and consumption. Like epilepsy, weakness of will is manifested intermittently, in 'fits', during which the agent's good principles are prevented from operating. What is common to sleep, weakness of will, and epilepsy, is that they all involve a temporary paralysis or inhibition of normal function.

14. 457a21-33 For the 'atrabilious', see Glossary, s.v. melancholikos

15. 457a33-b6 Aristotle here thinks of somnolence which risen,

as distinct from

full-blown

sleep,

develops when the hot matter has which

ensues

when

the hot matter

undergoes 'reflux'. It flows back downward ‘from where the heat fails' (b3), i.e. from the point where the hot matter can rise no further, and cooling begins to take place. The concept of 'reflux' will be picked up in the concluding summary (458a26-27), where the cause of sleep is said to be ‘the reverse flow of the solid matter’.

45706

ON SLEEP

133

Ross (44, 260) located the ‘material cause’ of sleep in the concentrated hot matter exhaled from food (a33—b1), thus making this passage fulfil the promise of 455b13-16 to give all four Aristotelian 'causes' for sleep. But there is no indication in the text that a 'material cause' for sleep is being identified, and Aristotle probably did not have his 'four cause' methodology in mind at this point. See note 8 on 455b13-34.

16. 457b6-26 The problem of this section is to explain why, if the 'causes' of sleep are hot,

sleep itself should be a cooling process.

The assumption underlying this

problem, aptly named 'the Transmission Model of Causality' (Mourelatos, 1— 16), is the principle that a cause (aitia) must itself possess a property of which it is the cause, or at least cannot possess the opposite of that property. This principle had played a major role in Plato's Phaedo 102a-106e (see Gallop (1975), 186, 212-13, 221). For the principle in Aristotle, see Metaph.

1032b30-31,

1034a25-30, and GC 323b24-324a24 with Williams’ note on

323b24, 121).

Its far-reaching influence upon later thought has been traced by

A.C. Lloyd.

|

Aristotle puts forward two tentative solutions to this problem before giving his own preferred solution at 457b26—458210.

b20.

As has been said: i.e. at 456b25. There is no need to suppose, with Beare, a reference to PA 653a10. Other internal cross-references within the De Somno occur at 456b9-11 and 456b17 (to 455b2—8) and at 458a15 (to 456b1). For their significance in regard to the composition of the DS, see Lowe,

283, with

290, n. 18. b25-26. And as the solid matter descends, it produces unconsciousness, and subsequently fantasy (phantasia): For 'unconsciousness' see Glossary, s.v. eknoia. ‘Subsequently’, i.e. at the stage where the separation of pure and impure blood, described at 458410--25, is occurring.

For the relation of this process to

dreams, cf. 461a25-30, b11—20, and see note 18 on 458a10—25.

17. 457b26-458a10 As

Lowe

has

noted

(282-3),

this

section,

which

reconsiders

a conclusion

already firmly reached at 456b24, and reaffirmed at 457a33—b6 on the basis of much supporting evidence, reads like second thoughts, attempting to deal with an objection to the original account. Aristotle gives an account of the brain in PA IL7. As the coldest part of the body, it functions as a coolant, maintaining the body at a mean temperature (PA 652b16—26), by counteracting the heat from the heart. Its role in causing sleep, by cooling hot matter and making it return downwards, is similarly described at PA

653a11-20,

with

a cross-reference

to the DS.

The

empirical

basis for

134

NOTES

458a10

Aristotle's view of the brain and its relation to the heart is discussed by G.ER. Lloyd, 224-3.

18. 458a10-25 This paragraph fulfils the initial promise (453b11~14) to examine the nature and cause of both sleeping and waking. Sleeping has now been fully explained. No independent account of the waking state needs to be given, since Aristotle has already covered this by treating sleep as its 'privation. To be awake is to be in the state in which senses and intellect are functional.

It still remains, however,

to explain the transition from sleeping to waking, i.e. ‘waking up’. The key factor is the separation of the rarer from the thicker blood. Perception, which has been inhibited during sleep owing to the intermingling of these kinds of blood, becomes possible again, as soon as the thinner blood once more predominates in the head. This account makes clear that it is owing to the thickening of the blood that the sensory apparatus becomes disabled in sleep. Aristotle will stress later (461a25, b17-19, b26-27) that in order for dreams to occur, there must not be an excessive amount of blood remaining in the senseorgans.

On the importance of rarefied blood for perception, see Wiesner, 266--

7. Cf. also note 39 on 461b7-30. 415-21. For the description of the heart and its relation to Aristotle's other accounts of this organ, see G.E.R. Lloyd, 227-8. 19. 458a25-32 This recapitulation harks back to the 'four cause' methodology of 455b13-16, and may have been composed or adjusted to suit that passage. See note 8 on 455b13—34

and Lowe,

285-6.

Note, especially, the emphasis

upon the ‘final

cause' and ‘necessity’ for sleep in the closing reference to ‘preservation’ of the animal. a30-31. Since it is not possible for an animal to exist, should the conditions that produce it not obtain: The 'conditions that produce it' are, for any animal, its exercise of the powers in which it realizes its nature, and thus is an animal of

the relevant kind.

It was argued at 455b16-28 that sleep is necessary for these

conditions to be realized. In terms of the doctrine of 'conditional necessity' invoked at 455b16-28, waking activity is the end, or ‘final cause’, which necessitates sleep. But since sleep and waking are opposites, the ‘final cause’ of sleep will turn out to be sleep's own opposite, and thus cannot be identified with its 'formal cause'. The thesis that sleep has a 'final cause' seems, therefore, to represent an earlier stage of Aristotle's thinking about it than is reflected in the bulk of the treatise. See Lowe, 286—7 with nn. 32—33.

458233

135 ON DREAMS Chapter One

20. 458a33—b3 b1.

With what part of the soul it is apparent: This question occupies the whole of ch. 1, and is finally answered at 459a21-22. ‘Part’ corresponds to no word in the Greek, but Aristotle is asking to which of our faculties dreams should be attributed. Are they the work of perception or of thought? For the options 'thinking' and 'perceptual' (D1—2), see Glossary, s.v. noetikon and aisthetikon.

b2-3.

It is with these alone that we know anything: With these words Aristotle justifies the initial presumption that a dream must be the work of either perception or thought. 'Know' should be understood broadly here to cover 'cognition' in general. Cf. Metaph. 1036a2-6. Aristotle does not suggest that dreaming is a mode of 'knowing' in any strong sense of that verb. At b2 a dream is called an ‘affection’. See Glossary, s.v. pathos. The

rest

of

the

chapter

will

show

that

dreaming

is

not,

in

fact,

to

be

straightforwardly attributed to either perception or thought, although intimately connected with both. Their respective roles in the explanation of dreams will emerge as Aristotle's theory unfolds.

21. 458b3-9 Aristotle first argues against attributing dreams to perception: (1) The use of any special sense (sight etc.), or of perception in general, consists in perceiving in the relevant mode (seeing etc.), (b3--4). (2) In sleep we cannot perceive in any mode (b7—9). Hence, (3) our awareness of dreams does not consist in the use of any special sense

(b9). (1) is supported by appeal to sight, hearing, and perception in general. The use of sight is seeing, and so on for the other senses. The principle is a conceptual truth. (2) invokes our inability to see anything while our eyes are closed. The extension of this point to the other senses (b7—8) is more questionable, since for them there is nothing corresponding to the closure of the eyes. Moreover, at 462a19—25 Aristotle himself notes that faint sensory awareness may occur during sleep. In general, however, our inability to perceive during sleep follows from his account of sleep in the De Somno. For at 454b25-26 sleep was defined as an incapacitation of the senses. The inability to see during sleep is due not just to the closure of the eyes, but to the total perceptual shut-down in which sleep consists. (3) The conclusion 'it is not with perception, anyway, that we perceive a dream' sounds paradoxical. One might think that 1f we 'perceive' dreams at all, it must

136

NOTES

458b4

be with perception that we do so. But ‘perceive’ (b9) should be taken in a wide sense to mean 'be aware of. Aristotle is arguing that our 'awareness' of dreams cannot be attributed to any of the special senses as these are exercised in ordinary waking perception. See Glossary, s.v. aisthanesthai, aisthesis. He will show later how the senses are involved in the explanation of dreams. For now, he simply argues that to dream is not to exercise any of them in the ordinary way. The argument assumes that dreams can occur only during sleep. This is implicit in Aristotle's word for 'dream' (see Glossary, s.v. enupnion), and is, for him, a conceptual truth.

b4-6.

Cf. 459a19-20, 462430--31, see Introduction, 20-1.

For the distinction between 'common' and 'special' objects of the senses, cf. DA 418a7-26, 425a14-b12. The 'common' objects are those perceivable by more than one sense. Besides those named here (shape, size and movement), they include rest, number, (DA 418a17-18, Sens. 43729), unity (DA 425a16), roughness and smoothness, sharpness and bluntness, (Sens. 442b4—7), and (possibly but more dubiously) time (DM 450a9-10, 451a16-17). The exact relation between these 'common sensibles' and the faculty of 'common sense' remains obscure, and was perhaps never completely worked out by Aristotle himself, although the De Somno sketches a view of the relation between the common sense and the special senses that substantially develops the position of the De Anima. See note 6 on DS 455a12-b2 above, with the exegesis of Kahn, 6—22, and Modrak, 66-8.

At DS 455a15-16 a ‘common faculty’ was said to be attendant upon all of the special senses. It is uncertain whether (as Ross supposed, p.259, note on 455a12-20) that faculty should be equated with the ‘common sense' that was said at DA 425227 to apprehend the 'common sensibles. But it seems clear, at

least in the essays on sleep and dreams, that all objects of perception, both those proper to single senses, and those common to more than one, must be Colours, sounds, and experienced in a unitary, 'common' consciousness. flavours are registered by a single sensory apparatus, not by a plurality of separate subjects. Likewise, it is one and the same sensorium that registers the shape, size and movement of an object. It can discriminate various kinds of data through its different modalities; it can be aware that it is doing so; it can also recognize that these data differ from one another, yet attribute them to a

single common source, the unitary physical object to which they all belong. ΑἹ] of these operations embrace ‘common’ and 'special' sensibles alike, apprehended ‘in the unifying and discriminating activity of a single centre’ (Kahn, 11). In the present argument, however, the distinction between special and It merely makes the point that common sensibles does no major work. perceiving 'in general’ (b4) includes common sensibles as well as objects proper to single senses. The argument that sensible properties cannot be perceived during sleep will therefore apply no less to shape, size and movement than to

458010

ON DREAMS

137

colour, sound and flavour. Aristotle mentions both kinds of object, presumably, because dreams contain data of both kinds In the example to be given just below (458b14) the dream figure is both 'pale' (the object of a special sense, vision) and 'approaching' (a common sensible, movement).

22. 458b10-15 This passage contains the first of two arguments against the first option canvassed at 458b1-2, i.e. that dreams are the work of 'thought' or (as the capacity is here called) judgment. For this capacity, see Glossary, s.v. dokein, doxa, doxazein. The argument runs as follows: (1) During sleep we say not only that (p) something approaching is a man or a horse, but also (g) that it is pale or beautiful (010-11). (2) We could not judge either that p or that q, whether truly or falsely, without perceiving the approaching item (b11-13). (3) During sleep we are not perceiving anything at all (as was argued at 458b7-9 above). Hence, (4) our saying that p or that g during sleep cannot be attributed to judgment. Hence, (5) our awareness of dreams cannot be attributed to judgment. In asserting that ‘we say' such things as ‘that is a man’ or ‘that is pale’, Aristotle does not mean that those words need be uttered by the dreamer aloud. The verb translated 'say' can also mean ‘think’, 'thinking' being regarded as a sort of inner speech, or 'saying to oneself.

Cf. 459a6, 460b21, 461b2, 461b24-462a8.

Aristotle has been taken to mean that the ascription of sensible qualities, such as paleness or beauty, requires perception, whereas identifications in terms of ‘substance’ concepts,

such as 'man' or ‘horse’ do not.

Thus

Siwek,

228, n. 7.

Ross (268) writes: "That dreaming is not an exercise of the faculty of opinion Aristotle infers from its containing a definitely sensuous element, such as the ascription of whiteness to the dream object'. It would,

however,

be

as implausible

to hold

that

one

could

identify

an

approaching object as a man or a horse without perceiving it, as to hold that one could ascribe beauty or paleness to it without doing so. And if 'none of those things' (b11—12) covers all of the examples just mentioned, than any assertion with respect to an object approaching will require perception of it. What makes it plausible to claim that perception is required for such assertions as 'that 1s a horse' or 'that is pale' is that the item is represented as 'the thing approaching' (b10, cf. b14). It is that phrase that carries the weight of the argument, rather than any distinction between attributions of substance and quality concepts. For an item described as 'approaching' is ipso facto picked out from one's perceptual field, and so perception is necessary if one is to identify it as a man or to describe it as pale. But perception is no less necessary for the former task than for the latter.

138

NOTES

458b15

What, then, justifies treating dreamers as making assertions about an item described as ‘the thing approaching? The answer is contained in b14—15, where it is argued that the soul, in effect, does make such assertions during sleep, because 'we judge that we see, equally, that the figure approaching is a man and that he is pale'. In dreaming, we often seem to be seeing something, and it is this aspect of the experience that cannot be attributed to pure judgment. It might be objected that not all dreams need be represented in such terms, and that some

cannot

be so described.

Thus,

Aristotle

never

considers

dreams

lacking any ‘sensuous clement’ at all, e.g. dreaming that 2 + 2 = 5. Nevertheless, since dreams are often characterized in terms of something we seem to see, it is fair to argue that judgment alone cannot account for dreams in general. Greek idiom treats a dream as something 'seen', and lacks the construction 'S dreamt that p', in which the dream is treated as a propositional object.

See Introduction, 7-8.

The claim that ‘we say’ such things as ‘that is a man’ or 'that is pale’ has been attacked by Malcolm as incompatible with the supposition that the subject is asleep. The matter remains controversial. See Introduction, 53-4.

23. 458b15-25 This passage contains a second argument against attributing dreams to ‘judgment’. It opens (b15-18) with an analogy between dreaming and waking perception. Just as we often engage in some thought about what we perceive, so in sleep 'we sometimes think other things over and above the appearances’ (b17-18). What we think in dreaming of X is distinct from our dream, just as what we think when perceiving X is distinct from our perception of it. We cannot, then, attribute our dream awareness to judgment, any more than we could attribute our waking perception to that faculty. Thus, our thought that the item appearing to us is a man or is pale should be distinguished from the appearance itself. Only the latter is the dream. As such, it is not the work of judgment, any more than are our waking perceptual experiences. Aristotle adds (b18-20) that 'this would be apparent to anyone who should make a mental effort to remember on arising’. That is, when one tries to remember a dream one has just had, one can distinguish the dream appearance from the identification of it as a man , or as pale, or whatever other judgment one may have made about it. The analogy with perception will play a vital role in Aristotle's account of dreaming later, especially at 461b24-30. But it is questionable in at least two ways. (1) It leads Aristotle to draw a sharper distinction between dreams and thoughts than we should be inclined to draw. See Introduction, 9. (2) Aristotle assumes that in dreaming we identify what we passively experience, much as we identify the content of waking experience. But it is arguable that we do not so much identify what is 'given' to us in dream experience as ordain what the items in that experience shall be.

See Introduction, 55-6.

458020

ON DREAMS

139

520-25. Aristotle now reinforces his argument against attributing dreams to judgment with an obscure example. He mentions dreams in which people think they are arranging a given set of items according to a mnemonic system. This system is mentioned in three other places (Top. 163b28—33, DA 427b18-20, DM 452a8-26). It is sometimes called the ‘Place System', and was widely practised in antiquity as a method of training the memory. See Sorabji, 22-31, and Yates, chs. 1—2, especially 45-50. Images of items to be remembered were superimposed upon a basic and unvarying set of 'place' images, e.g. the different rooms in a house, which provided a background against which the items could be placed, and retrieved later when needed. By this method, a person could retain and recall at will, for example, a set of points to be made in a speech, or a list of names to be recited. In the present passage the words ‘putting into its place before their eyes some other appearance (phantasma) apart from the dream’ (b23--24) refer to this technique. The example has occasioned various misunderstandings. One is that by 'putting into its place', Aristotle means that in subsequent recall of dream experience the subjects put one image 'in place of' another, i.e. put one image where another properly belongs. Thus Ross (267-8) speaks of people 'rationalizing' their dreams, and of those who 'substitute new objects for those originally presented', through the use of a mnemonic system. Aristotle would then be referring to what Freud called 'secondary elaboration’, i.e. the distortion of dreams upon waking up (cf. Devereux, 5, 131, n. 282, 137, 270). Any such interpretation must connect the use of the mnemonic system with an effort to memorize dreams while they are in progress and a subsequent incorrect recall of those same dreams upon waking up. But it is clear that Aristotle is no longer talking here, as he was at b18—20, about recall upon waking up. He is talking only about what happens while the mnemonic system is being used during the dream itself. Secondly, Aristotle has been taken to be referring here to a dream within a dream (Drossaart Lulofs

1947, xxviii).

The complex situation of telling dreams

within dreams, though once mentioned by Plato (Theaetetus 158c), has no relevance to Aristotle's argument here. It is true that if by ‘such dreams' (b20) Aristotle were to mean ‘dreams exactly like the experience just mentioned’, i.e. the experience of trying to remember a dream upon rising, then he would be referring to dreams in which people think they are trying to recall the experience of a previous dream. But he need not mean that at b20, and it would be a pointless complication of his example. It does not seem at all likely that people would dream of themselves recalling a previous dream with the help of a mnemonic device, let alone that they would be said to follow that procedure ‘often’ (022). A simpler reading is as follows. Aristotle is led by his own remarks about trying to remember dreams when we wake up (b18-20) to reflect upon efforts

140

NOTES

458b25

of memory that seem to be made occasionally during a dream itself. In ‘such dreams’ (020), people may think that they are recalling (or perhaps memorizing) with the help of a mnemonic system a certain set of items (not, as Ross assumes, the contents of the present dream). While doing this, they are having thoughts ‘apart from the dream’, i.e. distinguishable from the dream proper. For the images of their own system of places, the images of items that are being put into those places, and the mental activity of doing this, are not the dream proper, but are thoughts 'apart from the dream'. So, once again, the dream is not the work of judgment. Rather, 'it is what we are thinking that we cognize with judgment' (b25), whereas the dream proper must be attributed to some other faculty. Suppose, for example, that I dream of going to the market. I may think that I am using a system of 'places' to remember the things that I intend to buy. If I use an image of my study, in order to memorize, or to retrieve from memory, e.g. a bag of oranges, then neither the image of my study nor that of the oranges, nor my imposition of the one upon the other, is to be identified with the dream of myself going shopping. So, whereas all those things may be attributed to judgment, the dream itself should not be. We might, to be sure, have reservations about this argument. We might object that the images of the oranges and the study and the operations performed with those images, are themselves parts of the dream, or are not to be distinguished from it as sharply as Aristotle suggests. We might also ask why those mnemonic images and operations should be treated as the work of judgment, and why it should follow from their being the work of judgment that the dream itself cannot be. At most it would seem to follow that the dream need not be an object of judgment, but not that it cannot be such an object. ΑἹ] of this is debatable. But if we have approximated a grasp of Aristotle's meaning, the passage has nothing to do either with 'dreams within dreams' or with 'secondary elaboration’.

24. 458b25-29 Having reached an impasse over attributing dreams either to perception or to judgment, Aristotle now says that one thihg, anyhow, is clear: ‘it is the very same faculty by which we are deceived during illnesses, even when awake, that also produces the affection [i.e. being deceived] during sleep‘. He is suggesting a common origin for deception in dreams and the delusions of the sick. This key point will be developed more fully at 460b3-27. The factor common to both cases, as Schofield (124) says, is that 'as with the pathological seeing of aspects, so in dreaming things appear to be what they are not, and are often mistakenly taken to be what they appear. Things may also 'appear to be what they are not', even when we are in good health and are not all deceived as to how they really are. Thus, 'even to those who are in good health and who know otherwise, the sun still seems to be only

458b29

ON DREAMS

one foot across' (b28-29).

This example

141 was used at DA

428b3—5

to make

a

similar point, and will be used again below at 460b18-20, to show that false appearances are not necessarily endorsed by our judgment. Things may strike us in a certain way, even when we know quite well that they are not really that way. In dreams, by contrast, appearances usually go uncontradicted by judgment.

See note 40 on 461b30-462a8, and cf. Schofield, 116 with n. 37.

The sun example has an important bearing faculty referred to here (phantasia). No mental appearing to be one foot across, at least if an occur only in that object's absence. Cf.

upon the interpretation of the imagery is involved in the sun's image of an object is taken to Schofield, 121-2. See also

Introduction, 23-4, and Glossary, s.v. phantasia.

b29. The verb translated 'seems' (dokein) seems to be used loosely for ‘appears’ (phainetai), which occurs in the other two uses of the same example. Elsewhere it is used for 'thinking' or ‘judging’ something to be the case, as contrasted with what ‘appears’ to be the case. See Glossary s.v. dokein, phainesthai. Here, however, Aristotle cannot mean that people ‘judge’ the sun to be one foot across, since he says that they 'know' otherwise. 25. 458b 29—459a1 'But whether in fact the imagining part of the soul and the perceptual are the same or different': For 'the imagining part’, see Glossary, s.v. phantasia, phantasma, phantastikon, and Introduction, 22-5.

In setting aside the question whether the imagining part of the soul is the same as or different from the perceptual, Aristotle is assuming that the former is indeed the faculty referred to in the last sentence but one (458b25—28), whereby we are deluded when sick or when experiencing perceptual illusions. Aristotle thinks that the imagining and perceptual parts are, in some sense, 'the same' (see note 28 on 459a14—22). But his present point is that even if this were not so, the attribution of dreams to the imagining part still seems open to the difficulty raised initially (458b3-9) about attribution of dreams to perception. For even our experience of false appearances seems to require that we should be perceiving something, as we are when we misperceive things while awake. So the ascription of dreams to the imagination encounters that objection, even if the latter faculty is not identified with perception. The objection will be tentatively answered at 459a1—8 below. See next note. 26. 459a1-8 The tentative nature of the answer outlined here is marked by casting it in the form of a question. It will be developed in further detail later, and put forward as the correct solution. Cf. 460b22-27, 461b26—29. Aristotle suggests that sight and the other senses, although not affected by any external object, may nevertheless be affected somehow, just as they are for a waking person, though not in the same way. Each such affection of the senses ‘impinges in some way

142

NOTES

459a6

upon perception’, in that we are aware of something as a result of the senses being affected by internal movements (to be described in ch. 2). The affection of the special senses has an impact, in turn, upon the primary sense-faculty. Thus, ‘impinges in some way upon perception' (a4—5) might be glossed here as ‘somehow reaches consciousness’. a6-8.

And does judgment sometimes declare it an illusion ...?: (a6-8). This suggestion uses the distinction illustrated in the example of the sun (458b28-29) between 'appearance' and ‘judgment’. Sometimes judgment rejects an appearance as false. At other times it ‘follows along’ with the appearance, i.e. accepts it in an uncritical way. This foreshadows Aristotle's later account of dreams: occasionally the subject becomes aware of his sleeping state, and judgment then rejects the dream appearance as false (462a2-8). But in general, the appearance goes uncorrected by judgment, and thus prevails by default.

27. 459a8-14 Aristotle now argues that dreaming, despite its not being the work of ordinary perception, must nevertheless be attributed to the perceptual part, seeing that it belongs to the same part as does sleep. 410.

For then it would be unqualified seeing and hearing: (410). For this construal of the Greek, see textual note 13. With this translation, Aristotle is saying that if dreaming were an ordinary exercise of the senses, it would actually be seeing and hearing of the ordinary kind. Beare gives, ‘it would be possible to hear and see in the simple sense’: i.e. if dreaming were the work of ordinary perception, it would be possible to see and hear in the ordinary way while asleep. On either view, Aristotle is glancing back to his earlier argument (458b7—9), that ordinary seeing and hearing during sleep are not possible.

all.

But just how and in what manner is what has to be examined: Aristotle is not yet addressing the question of "what a dream is and how it occurs', as Beare and Hett suppose. For that question will not be raised till the start of ch. 2 (459a23-24). The present line is better understood as follows: given, as has just been stated, that dreaming is not an exercise of perception of the ordinary kind,

in what way is it an exercise of perception? For that is the question that Aristotle proceeds to answer. He goes on (459a11—22) to hypothesize that while dreaming is, indeed, the work of the perceptual part, it is not an ordinary exercise of that part, but an exercise of it in its 'imagining' (phantastikon) capacity. 811-14.

The (1) (2) (3)

argument runs as follows: Sleep belongs to the perceptual part. Dreaming belongs to the same part as sleep. Hence Dreaming belongs to the perceptual part.

459a14

ON DREAMS

143

(1) has been established by the argument of DS 453b24—455b13. (2) is asserted at a13-14 without argument. Aristotle presumably thinks it self-evident, because he takes dreams to be linked conceptually with sleep (459a19-20, 462a30-31). See Glossary, s.v. enupnion. 28. 459a14-22 This difficult passage contains the answer to Aristotle's initial question (458a33—b2) as to the faculty to which dreaming should be attributed. It is the work of the perceptual part, but belongs to it in its ‘imagining’ (phantastikon) capacity. For a general discussion of the role of phantasia in dreams, see Introduction, 21-8.

At a14—15 reference is made to the discussion of phantasia in DA III. ch. 3. The argument runs as follows: (1) The imagining part (to phantastikon) is the same as the perceptual, yet they differ in their 'being' (315-17). (2) Imagination (phantasia) is the 'movement produced by a sense in the course of its active functioning' (317-18). (3) A dream is a kind of 'appearance' (phantasma) (a18—20). Hence, (4) dreaming must be the work of the perceptual part, but belongs to it in its imagining (phantastikon) capacity (421-22). (1) In saying that the imagining part is 'the same' as the perceiving, yet that they differ in their 'being', Aristotle probably means that they have the same physiological basis, yet one would have to give different accounts of what each of them is.

See also Introduction, 22-3.

(2) repeats a formula for phantasia given in DA 429a1-2. It would explain mental images in terms of 'traces' left in the sense-organs from actual perceptual experience, or in Hobbes's well-known phrase, 'decaying sense’. Aristotle will enlarge upon this theory in chs. 2-3. See also Introduction, p. 20, 23. (3) appeals to ordinary usage. Since a phantasma in sleep is 'what we call' a dream, we are justified in ascribing dreaming to phantasia. a20.

whether it occurs simply or in some particular way: (a20). With these parenthetical words Aristotle puts aside the question whether it is just any kind of phantasma in sleep that we call a dream, or whether we confine the word to phantasmata generated in a particular way. The latter is his own view. He has already argued that not every phantasma in sleep is a dream (458b24—25), and he will insist later (462a15-28) that other kinds of phantasma in sleep are not properly called ‘dreams’. For present purposes, however, this point is immaterial. To show that a dream is the work of imagination, it is enough that it should be some kind of phantasma in sleep: just what kind does not, for the moment,

matter.

144

|

NOTES

459a23

Chapter Two

29. 459a23-28 Aristotle now enlarges upon the suggestion that dreams are due to imagination (phantasia), understood as ‘decaying sense’. Much of this chapter is devoted to showing how this can happen.

30. 459a28-b7 This passage provides a model for understanding the persistence of change when its initial source is no longer operative. Aristotle is thinking of a mover's power to impart movement to other things indirectly through a series of intermediate movers.

Cf. Phys.

215a14—17, 266b30,ff.

a28-33. Objects moving in space: Aristotle's verb is sometimes rendered ‘projectiles’. It can cover any form of local movement, including missiles borne through the air. But the general point could:just as well be illustrated by the effects of a pebble in water as by the disturbance of air. Both media are mentioned at a32-33. a32. Movement is said to continue ‘until a standstill is reached’. This must mean that a chain reaction continues until the impetus given by the original mover is exhausted, so that all parts of the air or water are once more at rest.

According

to Beare, Aristotle is thinking of successive displacements, ‘until the last thing (portion of air) has come into the place of the first movement’. But this would be irrelevant to his main point. He simply needs to observe that movements may continue, even when the stimulus giving rise to them is no longer operative. That is all he requires to establish that dream images can occur after the perceptual stimuli giving rise to them have ceased. b1-2.

b3.

The general point is now applied to qualitative change, e.g. heating. This extension is relevant, since Aristotle views perception as itself ‘a sort of qualitative change. See Glossary, s.v. kinesis, alloiosis.

Until the starting-point is reached: In what sense is it meant that the series of changes ends with 'the starting-point’? This is hard to understand in the case of heat, which is naturally thought of as dispersed, not as returning to its point of origin. Beare takes the phrase to refer to successive displacements of 'the cold’, until the whole object has been affected by the heat transmitted from one part of it to another. But 'the heating of the whole’ does not correspond in any obvious way to the ‘standstill’ reached in locomotion. Nor is there anything analogous to this in the case of perception, which the present example is supposed to illuminate. Cf. Ross (271) on 459b3. The general sense required in all three cases — locomotion, transmission of heat, and perceptual awareness —is that an initial moving agent can produce

459b7

ON DREAMS

145

indirect effects, through successive intermediate agents, even after its own movement has ceased. The 'domino' effect will persist until the impetus from the original movement has been exhausted. Thus, when heat is transmitted from a hot object to a sentient creature, it will travel through a series of contiguous regions of the body until it reaches the central sense-organ, the heart. This process provides a model for understanding what occurs in perception more generally. The 'starting-point' is therefore perhaps best understood here as referring to the heart. The term is often used in that sense elsewhere.

b7.

See Glossary, s.v. arche.

Both in depth and on the surface: i.e. both in the central sense-organ and in the surface organ, e.g. the eye.

31. 459b7-23 In this section Aristotle offers empirical confirmation of his 'decaying sense’ theory. Thus he supports it not merely by theorizing about on-going movement and change in physical objects generally (as in the preceding paragraphs), but also by appealing to familiar kinds of after-effect that attend perception in particular. The range and detail of his examples illustrate his acute powers of observation.

32. 459b23-460a23 This notorious passage, together with 460a23-32, has been rightly suspected (most recently by Modrak, 209, n. 50) of being an interpolation. The supposed mirror phenomenon is included by the elder Pliny among many ancient superstitions associated with menstruation (Natural History, VII. 64-66). It is hard to believe that Aristotle gave any credence to such old wives’ tales. Nothing of the sort is ever suggested in the scientific accounts of menstruation given in his zoological treatises (GA I. 19, HA VII. 2). Whoever wrote the present passage, it is of disproportionate length, and remains poorly integrated into the text. Note, especially, the awkward transition to the main issue at

460232. Even if the supposed mirror phenomenon were genuine, its relevance to the argument would be hard to understand. The passage has, nevertheless, been defended as authentic by various commentators xxxiii, Ross, 272, Siwek, 231, Preus).

(Drossaart Lulofs

1947,

xxx—

b23. what happens with mirrors: It should be borne in mind that ‘mirrors’ in antiquity were made of polished metal. Cf. 'bronze' at 460a14. The mirror phenomenon is said to show that (1) the sense-organs are rapidly sensitive, (2) even to slight differences; and that (3) the organ of sight is not only affected by, but also acts reciprocally upon, its object. These three points are repeated at 460a23, and are said to have been made 'plain'. But none of them seems at all clearly connected with the prime requirement of Aristotle's

146

NOTES

460a8

‘decaying sense’ theory, that the impact of perception should continue to be experienced after its original source has departed. That is the essential point needed for his account of dreams. It is resumed once again at 460a32-b3 after the digression. Yet it would hardly have been reinforced by the mirror phenomenon, even if the latter were genuine. Of the three points just noted, (3) would have some relevance to the main argument, if it showed that the eye is not merely a passive receptor, but also an agent that can produce effects in its own right. But one may doubt whether the mirror phenomenon would provide any evidence for this, and whether any inference could be drawn for ‘the sense-organs' (b23-24) in general from this particular case. Possibly, as suggested by Sprague (1985), an analogy is intended between the reflecting mirror and the seeing eye, which is itself a shining object (a2-3). Just as the mirror not only reflects the eye of the menstruous woman but also retains traces of its impact, which may be hard to eradicate, so an ordinary eye not only sees its object but may also retain traces of the relevant sense-impressions, persisting long after interaction between eye and object has ceased. This would give point to stressing the difficulty of getting the stain off the mirror (b31-32, 2312-23), and it would link the phenomenon with the main point required by Aristotle's 'decaying sense' theory. But it would not support any of the three points which the phenomenon is alleged to make 'plain'. a8-9.

For the nature of semen and of the menses is the same: Semen and menstrual blood have a common nature in that both are 'concocted residues’.

For elaboration of this theory, see GA 726a6,ff.

The present words explain the

existence of a difference in the eyes of a menstruous woman

(Ross, 273, Siwek,

232, n. 35), not the fact that it is not apparent to us (Beare). The reason why the difference is not apparent has already been given in the words 'they are full of blood vessels by nature’ (a5-6). The present words probably explain the difference by suggesting that the bloodshot state of the eyes is 'akin to that produced in men by the emission of seed though less readily observable' (Ross). In both cases an organ has present within it dischargeable material of similar composition.

33. 460a23-32 Further evidence is here adduced from wines and ointments. The effects noted, unlike the mirror phenomenon, are genuine. But they provide no more support than does the mirror phenomenon for any of the points noted about the senseorgans at 323-26. At best they provide an illustration of how the effect of X upon Y can persist, even when X is no longer in contact with Y. They also give some support to the interpretation of the mirror passage proposed by Sprague (see previous note). For, as Barnes has suggested, perhaps ‘what Aristotle has in mind is that the wine or oil is to a distant herb as the mirror is to the distant

460432

ON DREAMS

147

eye: both are quickly (and persistently) affected by small differences’ (quoted by Sprague, 325, n. 1). 34. 460a32-b16 Aristotle here returns to his main theme, reaffirming the 'decaying sense' theory as 'plain from what has been said'. This is the first assumption needed for what follows. A second crucial point follows immediately, which contains Aristotle's key insight into the role of the imagination (phantasia) in dreaming. It develops the connection between dreams and perceptual errors already made at 459b28—-29. We are, Aristotle now points out, especially prone to error when emotionally aroused (b3—1 D), particularly to that of mistaking one thing for another in virtue of some slight resemblance between them. Severe illness can have the same effect (b11-16). What is common to such cases is a wrong assimilation of some perceived X to some other thing Y. The assimilation goes uncorrected by judgment, if its operation 15 impaired by emotion or by certain pathological states. This point will be applied in ch. 3, where Aristotle seeks to show that we are in a comparable state when dreaming. In sleep, he suggests, we may make a wrong assimilation, by mistaking a remnant from waking perception for an original sense-impression, and thus think we are perceiving a real exterpal object. See 461a14-b7, 461b21-462a8. The mistake goes uncorrected by judgment, because that faculty is usually suspended or impaired during sleep. Cf. Schofield,

122-3, and Modrak

(94,

101,

137-8), who

takes the absent co-

ordinating faculty to be 'common sense'. b4, b11-15. For the word rendered ‘emotional states' or 'condition', see Glossary s.v. pathos. Although dreaming is not 'pathological', it is akin to such states in a way that will become clearer in ch. 3. b15-16. They actually move themselves in accordance with the appearances: The patients move their bodies, e.g. shrinking in terror, according to the supposed movements of the animals in their fancy.

35. 460b16-27 Here Aristotle explains the nature of the conflict between 'appearance' (phantasia) and judgment, with reference to cases where the former is contradicted by the latter. For the example 428b2-4, and see note 24 on 458b25-29.

of the sun, cf. 458b28-29,

DA

b19-20. Frequently something else contradicts the appearance (phantasia ): Either our judgment contradicts the appearance that the sun is only one foot across, or it contradicts the imagining 'faculty'. Whichever way phantasia is understood here, the example suggests that what is contradicted cannot be a mental image, but must be an impression of the sun's size, which persists

148

NOTES

460b20

despite our judgment to the contrary (cf. 458b28). n. 51, and Introduction, 23-4. b20-22.

For the tactual illusion,

959a15-19, 965a36-39.

cf. Metaph.

See Schofield,

1011a33-34,

118, 135,

[ Probl.] 958b14-16,

It works well if the tip of the nose is rubbed between

the crossed index and middle fingers. The phenomenon was discussed by Croom Robertson and others cited by Ross (273-4), and has been investigated experimentally by Benedetti. The illusion, like the sun example, makes no appeal to mental imagery. For, unlike mental imagery, it can occur only in the presence of its object. b22.

Aristotle goes too far in saying that ‘if touch stood alone, we should actually judge the single object to be two'. It is not merely sight that corrects the tactual illusion. Rather, we allow for the fact that we are touching the object in an abnormal

way.

It is true, however, that the illusion is less easily obtained when

the object is visible. b22-27. These lines give a general explanation of illusions by a principle that will later be applied to dreams. 'Appearances' may result not only from 'movement' imparted to a sense-organ through its interaction with objects perceived, but also from 'movement' of the sense, should it be affected as it would be by an external object, but for some other reason. Thus, land seems to be moving for observers sailing past it. Their vision is moved 'by something else' (b27), i.e. because of the movement of the ship. Hence, their experience is like that of seeing a moving object in the ordinary way. At b26 Aristotle says that the land 'seems' (dokei) to be moving. Perhaps he uses the word loosely, as at 458b29, for 'appears' (phainetai). For this is not a case in which observers are liable to think that things are as they appear to be. See Glossary, s.v. phainesthai,

dokein. The sailing example illustrates Aristotle's point in a rather misleading way. In this case the perceived object (i.e. the land) appears to be 'moving' (in the English sense of 'move'), when it is really the sense, or its organ, the eye, that is 'moved' as a result of the movement of the ship.

‘movement’ possess.

But this case is special, in that

happens to be the property that the perceived object appears to

This is not, of course, a feature of all illusions, nor is the ‘movement

of an organ, in the way that an eye is 'moved' by the ship, the essential point in the proposed explanation of illusions. The key point is that a sense-organ may be affected just as it is by an external object having certain properties, but from some other cause. Given such a stimulus, an illusion may be created. That is the principle needed to explain dreams later (461b26-29). Cf. also 459a1—5, and see Glossary, s.v. alloiosis, kinesis.

460b28

ON DREAMS

149

Chapter Three

36. 460b28-461a8 The physiology of dreaming is now described in more detail. Aristotle begins by suggesting that we are aware of processes at night that go unnoticed by day because of stronger competing stimuli from the external world. Cf. 463a7—21,

464a6-17. a5-6.

Because of the reversed flow of heat from the outer parts to the interior: For this process, cf. DS ch. 3.

a6-7.

The starting point of perception: i.e. the heart.

37. 461a8-25 This important section compares the residual movements from perception with those of flowing water.

Whether we dream, and what sort of dreams we have,

will depend upon the nature of these movements, just as the existence and the content of reflections will depend upon the movements of the water. For the comparison of dreams with reflections, cf. 464b5-16,

and see Introduction,

16,

55. In terms of this model Aristotle can explain (1) why dreams sometimes do not occur at all (317-21); (2) why they are sometimes experienced as grotesque distortions of waking experience (a21-25); and (3) why they are often sufficiently like waking experience to make us think we are perceiving something real (461a25-b7). He can thus claim to have answered the question posed initially (453b17—18), ‘from what cause sleepers sometimes dream but sometimes not'. 222-23.

For 'atrabilious', see Glossary, s.v. melancholikos.

38. 461a25-b7 a27.

Makes the dreams coherent:

Cf. the dreams are not coherent' (461a22).

The

translation 'coherent' is based upon an emended text at both places (see textual note 17 and Ross, 277).

‘Coherence’ has often been suggested as a criterion whereby to distinguish dreaming from waking experience. Cf. Malcolm, 101-13. But Aristotle is not here concerned to provide such a criterion. If he did use a word meaning 'coherent' in this passage, it was in order to contrast the more ordinary dreams of a25-30 with the wild ones of 461a21-24. Some dreams possess a degree of internal coherence akin to stretches of waking experience. It is these that he thinks of as effects of sense-remnants that remain undistorted by internal disturbance, and thus 'replay' waking experience in something like its original form. Cf. Glossary, s.v. euthuoneiria. Such dreams are liable to give rise to mistaken judgments, because they simulate the conditions of normal perception. This is further explained at a27b7, where the point made at 460b22—27 is applied (cf. also 459a3-8). Even

150

NOTES

461b5

when a sense-organ has not been stimulated by an external object, there may occur a series of internal movements akin to those that result from external stimulus. When these movements reach ‘the starting-point’ (431), i.e. the heart, there occurs an experience akin to normal perception: ‘something is made to appear, and because of effects carried inwards from the organ of vision one judges that one is seeing etc’ (a27-29). b5-6.

In every case, then, something appears, yet what appears is not in every case judged to be real: This repeats, in a pithy way, the essential point of contrast between 'appearance' (phantasia) and 'judgment' (doxa), upon which Aristotle's argument continually turns. Cf. 460b16—22. See note 40 on 461b30-462a8, and Glossary, s.v. dokein, phainesthai, phantasia.

39. 461b7-30 This section enlarges further upon the physiology of dreams (b7-21) and moves towards a final solution (b21—30). b7-11. A key point made earlier (460a32-b16) is picked up and applied to dreams: just as we are most easily deceived when in the grip of emotion or severe illness, so in sleep we are liable to make mistaken assimilations, which usually go uncorrected by judgment. b11-21. The physical processes described here are hard to understand in detail. The movements of sense-remnants are carried 1n the blood as it moves towards the heart during sleep. Cf. 458a10-25. These movements are, at some point, 'released', rise to the surface, and continue to be active in a smaller amount of blood remaining in the sense-organs (b17-19). Through their activity, they cause the same sort of perceptual awareness as does an external item 1n ordinary perception. For the stimulus they give to the sense-organ is continuous enough to give the illusion of normal perception. It is with this continuity in mind (cf. 'continuously' at 461a10) that Aristotle compares the behaviour of these movements, as they return to the sense-organ, with that of artificial (?toy) frogs, 'that float upwards in water as the salt dissolves' (b15-16). Two explanations of the frogs' movements have been suggested (Siwek, 234, n. 62). (1) They have, initially, a specific gravity greater than that of water, and so remain submerged at the bottom of their vessel until salt is added to the water. As the salt dissolves, their specific gravity decreases, relative to that of the water, until they are light enough to rise to the surface. Alternatively, (2) the frogs have a lower specific gravity than that of water, but are initially fixed in a bed of salt at the bottom of the vessel.

As this

salt dissolves, they are released from it, and so rise to the surface. On either explanation, the frog comparison will illustrate the continuous character of the movements stimulating the sense-organs. Just as the frogs keep on rising to the surface, one after another, so a continuous series of stimuli to

461012

ON DREAMS

15]

the sense-organs occurs: as one movement perishes another will take its place (b14—15). This series of movements enables dream awareness to be readily assimilated to normal perception. b12-13. Some potentially, but some actually: Aristotle thinks of some of the movements as actually occurring within the blood, whereas others are merely present in the blood 'potentially' i.e. are capable of being reactivated (Ross,

277).

b19-21.

|

The

movements

are described

as

‘taking

on

a resemblance,

as cloud-

formations do, which people liken now to men and now to centaurs, as they change rapidly’. Aristotle is thinking of rapid changes in the appearance of shifting cloud—formations, causing us to assimilate them to a variety of different things in quick succession. Similarly, the continual ‘shifting’ of movements in the blood (b18) would account for the ‘fluid’ character of dream experience. The cloud comparison also points to the importance, for Aristotle's concept of phantasia, of interpreting what we perceive in terms of a gestalt taken from something quite different. We assimilate a cloud to a man or a centaur in virtue of only a ‘slight resemblance’ between them.

Cf. 460b6, b8, b12.

b21-30. Aristotle finally applies the foregoing physiological story to show how dreams involve a kind of misidentification. This important but difficult passage contains a textual problem, for which see textual note 20. The translation is based upon Ross's text. For the technical term 'sense-impression' see Glossary, s.v. aisthema, and Sorabji, 82-3 (on DM 450a31-32). The revised interpretation which follows is due to remarks by Woods, 187.

Aristotle says that a sense-remnant which persists when its real original has departed ‘is like Coriscus, even though it is not Coriscus' (b23-24): the senseremnant is like, though not identical with, an ordinary sense-impression of Coriscus. By contrast, in normal waking perception, there is no assimilation of a sense-remnant to a sense-impression, hence nothing gets misidentified. The subject has a genuine sense-impression of Coriscus, and correctly attributes it to a real external object (‘the actual person out there’, b25-26). In dreaming, on the other hand, the judging part of the soul ‘is moved by the movements in the sense-organs, as if it were perceiving’ (b27-29).

Consequently, ‘what is like something is judged to be that very thing' (029). Thus a mere sense-remnant is mistaken for a sense-impression which it resembles. Hence the subject wrongly believes himself to be perceiving a real external object. And 'the effect of sleep is sufficient to make this escape notice'

(b29-30). With this analysis we reach the heart of Aristotle's explanation of dreams. They involve the mistaking of a mere sense-remnant for a genuine senseimpression, which leads the subject to suppose that a real object is before him. The experience of a sense-remnant in sleep is taken to be an experience of

152

NOTES actually seeing Coriscus, identical with, the latter. like something is judged in dreaming only for so

461b26

when the former experience is merely like, but is not This is the confusion referred to in the words 'what is to be that very thing’ (029). This confusion will occur long as the subject remains unaware of being asleep.

As soon as he becomes aware of this, the confusion is instantly dispelled.

b26-27. Unless it is completely held in check by the blood: The point of this proviso is unclear. Does it specify a condition that inhibits the judging part from identifying a perceived object as Coriscus? Or a condition that prevents the simulation of ordinary perception from occurring? The latter interpretation is simpler. If the blood remaining in the sense—organs is excessive in amount, or perhaps impure in quality (cf. 458a10—25), the residual movements will not be able to produce the illusion of sensory awareness in which dreaming consists. For a more complex interpretation, see Ross 278.

40. 461b30—462a8 Once again Aristotle explains the conditions under which appearances are not accepted by judgment. Cf. 461b5-7. In the example of double vision used here, the 'appearance' (phantasia) (a8) that gets contradicted cannot be a mental image, but must be the appearance of one object as two caused by pressure of a finger beneath the eye. Aristotle suggests that what prevents the appearance from being endorsed by Judgment in this case is the subject's awareness of the pressure of the finger, i.e.

awareness of the cause of the appearance.

This is more plausible than his

earlier suggestion at 460b20—22, in connection with the counterpart illusion of touch, that what appears to one sense is contradicted by the report of another. Such an explanation would not cover the present case, where the appearance is contradicted by sight itself, once the factor causing double vision is removed. This account is closely parallel to that of the role of judgment during sleep, in cases where it refuses to accept dream appearances. It refuses, not because the appearance is contradicted by another sense, but because the subject becomes aware of being asleep, and is thus aware of the special conditions that cause the experience. For the importance of recognizing that one is (or has been) dreaming, see Introduction, 27, 54.

41. 462a8-15 This passage contains the earliest description of 'hypnopompic' and 'hypnagogic visions that are occasionally experienced between sleeping and waking. They are introduced to support the theory that 'apparitional movements' (Kineseis phantastikai), i.e. movements producing appearances (phantasmata), occur in the sense-organs. Upon waking, Aristotle argues, it is sometimes possible to detect these visions as movements present in the organs. He does not explain how they are detected. But he takes them as evidence for movements subsisting within the eyes, like the frightening objects seen in the dark by children, which

462a15

ON DREAMS

153

they vainly try to shut out by covering their heads. Compare, also, the afterimage phenomena adduced at 459b5-18. All such phenomena, Aristotle would argue, must be due to movements within the eyes, since there is no current external stimulus to cause vision.

42. 462a15-31 Aristotle now concludes that a dream is ‘a certain sort of appearance (phantasma), and specifically, one that occurs during sleep' (a15-16). Here he approaches his formal definition of a dream, to be reached finally at a29—31. He excludes (al6-17) the visions of youngsters just mentioned. Because these occur wbile their subjects are awake, they do not count as dreams. Nor does ‘any other that may appear once the senses have been set free’ (al7—18). Waking is here treated as the freeing of the senses from the 'fettering' of sleep (cf. 454a32—b1, 454b10—11, b25-27).

Once it has occurred, the subject can no

longer strictly be said to be dreaming.

It is part of the concept of a dream, for

Aristotle,

sleep.

that it should

occur during

See

note 27 on 459a8-14,

and

Introduction, 20-1, with Glossary, s.v. enupnion. At a18-29 the definition of a dream is narrowed down still further. It is not even the case that every appearance in sleep is a dream. Thus Aristotle excludes: (a) faint perceptions from external stimuli (a19—25); (b) cases where sleepers answer questions put to them (a25—27); and (c) true thoughts occurring over and above the appearances (phantasmata) (a28-29). (a) If someone is aware of a barking dog while asleep, this cannot count as a dream of a barking dog. For it is not a mere sense-remnant that is being perceived, but a real (1f muffled) sense-impression from an external source.

(b) Aristotle gives no examples of questions and answers. But perhaps they would include questions and answers regarding phenomena perceived under (a). Sleepers may not only faintly hear a dog barking, but may even be able to answer, correctly or incorrectly, when asked what they hear. Their identification of the object should not be counted as dreaming, any more than the identification of something perceived while awake should be counted as perceiving (cf. 458b15-18). (c) Thoughts ‘over and above the appearances (phantasmata) (a29) have already been distinguished from dreams by the argument of 458b15-25. Modrak (215, n. 24) has interpreted the 'true thoughts' of the present passage as ones that 'contravene' the dream imagery, i.e. those that reject the appearances as illusory. She also (101) cites 458b10-459a22 as showing that the conflict between thoughts and appearances (phantasmata) establish the latter as 'representational' entities, possessed of a propositional structure and a truthvalue. By 'true thoughts' in the present passage Aristotle may well mean thoughts that one's current experience is merely illusory (cf. 462a2—7); and he has repeatedly invoked cases where 'appearances' are contradicted by thought or judgment. But we should beware of translating the present line 'true thoughts

154

NOTES

462a15

contrary to (para) the appearances'. For in 458b15-b18, where the point made here is anticipated, para must mean ‘alongside’, ‘over and above' (cf. 458623). Aristotle's interest in distinguishing thoughts from dream appearances is not limited to cases where they are in conflict. At 458b10—25 he has argued, quite generally, that dream appearances, like their counterparts in waking perception, should not be attributed to the cognitive faculty (doxa) whereby we make judgments of any kind about them, whether true or false (b12-13). By this argument, the dream appearance is as distinct from a false judgment based upon it as it is from a true judgment contradicting it. It remains doubtful whether the appearance itself possesses a 'truth-value' of the sort requiring 'propositional structure’. For the sense in which Aristotle is prepared to call dreams ‘false’, see Introduction, 18-19. The various 'appearances' excluded here show how broadly Aristotle can use the term 'appearance' (phantasma). None of the items here distinguished from dreams could plausibly be regarded as mental images. See Glossary, s.v. phantasia, phantasma. The exclusions under (a) follow from (1) Aristotle's restriction of dreams to what occurs in sleep (a15-16), together with (2) his definition of sleep as perceptual incapacitation (454b9-11). It follows from (2) that any perception of sounds, smells etc. is incompatible with the supposition that the subjects are fully asleep (cf. 458b7-9). Hence, given (1), they cannot be dreaming. Aristotle's reluctance to allow that, strictly, perception can occur during sleep is shown when he says that the subjects perceive only 'in a certain manner' (a19—

20). He shows similar reluctance in connection with the subjects who answer questions put to them during sleep, by saying that ‘in the case of sleeping and waking, when one of these states is present in the ordinary way, it is possible for other to be present in a certain manner’ (a26-27). This forestalls a possible objection that, given Aristotle's definition of sleep, it is hard to understand how we can have any awareness in sleep at all. Aristotle might explain the phenomena of somnambulism along such lines (GA 779a12-19, cf. DS 456a2429). Despite being asleep, sleep-walkers might be said to be awake 'in a certain

manner. dreams.

But it is doubtful that he could consistently apply this solution to See Introduction, 54.

It is a Freudian misreading of Aristotle to suppose that in this paragraph he is considering the type of dream in which external stimuli become woven into its content

in altered

form.

He

does,

however,

notice

later (463a11—17)

that

internal organic movements may be misinterpreted during sleep. He does not call these phenomena 'dreams', and they would be excluded, perhaps unfairly, by his present restriction of dreams to the effects of prior perception of external objects. See Introduction, 40, n. 78.

462a29

ON DREAMS

155

a29-31. Aristotle here reaches his formal definition of a dream: (1) it is ‘an appearance (phantasma) that arises from the movement of the senseimpressions’; (2) it occurs ‘while one is in the sleeping state’; and (3) it occurs ‘in virtue of one's being asleep. Each of these points is derived from the preceding analysis: (1) confines dreams to experiences derived from the aftermath of waking perception; (2) specifies that they must occur during sleep, unlike the waking visions and thoughts excluded at al6-18; (3) confines them to experiences attributable to sleep, unlike the faint sense-impressions and thoughts excluded at a19—29. As suggested above, (1) is perhaps too restrictive. But Aristotle can fairly claim to have analyzed the ordinary Greek concept of a dream, purged it of its religious associations, and reinforced it with a scientific theory.

43. 462a31-b11 This paragraph is an addendum that has not been integrated into the main treatise. Non-dreamers and late dreamers are mentioned also at HA 537b16—21 (for which see Appendix) Aristotle does not here consider whether their supposed failure to dream is genuine or whether they may be presumed to have had dreams and forgotten them. For the latter possibility, entertained once by Aristotle himself (453b18—20), see Introduction, 35-6.

156

NOTES ON DIVINATION THROUGH

462b12 SLEEP

Chapter One 44, 462b12-26 This paragraph contains introductory arguments for and against belief in divination from dreams. The opening sentence sounds non-committal, but Aristotle will take a more sceptical view of divination than the words ‘it is not easy either to despise it or to believe in it’ (b13-14) might suggest. In particular, he will not endorse the popular belief that dreams have ‘some significance’ (b15), to the point of allowing that al/ dreams afford a basis for divination. In granting that divination from dreams makes sense ‘as regards some matters’ (b16), he is probably thinking of the dreams that he views as ‘signs’ or as 'causes' of subsequent events at 462b26-463a31. But he would not accept the extension, suggested at b17—18, from those cases to other dreams. Most dreams that are ostensibly fulfilled in later events he will put down to ‘coincidence’ (463a31-b1). The arguments against belief in divination (b18—26) illustrate Aristotle's approach to the whole subject. Absence of rational explanation, he thinks, makes for disbelief. Yet no such explanation is to be found, either by attributing veridical dreams to God or by supposing that some people have special precognitive powers. For the argument against a divine provenance for dreams, cf. 463b15-18, 464a19-24, and see note 49 on 463b12-22. Aristotle is here rejecting a traditional view. See Introduction, 7-10. The idea that certain subjects have precognitive powers will be explored more fully below. Cf. 463b15-22, 464a17-b5. Aristotle will attribute such 'powers' to the subjects' nature rather than to skill (techne). Indeed, he would regard the interpretation of dreams as a genuine skill, only to the extent that it has a rational basis, as in the cases to be discussed at 462b26-463a31.

b12.

The divination that takes place during periods of sleep: With ‘divination’, ‘the art of has to be understood. See Glossary, s.v. mantike. The preposition en (‘during’) may suggest that this art is being practised while the subject is asleep. But Aristotle sometimes uses the preposition kata, which has been translated 'through' in the title of the treatise and which suggests, rather, that divination is

based upon dreams, inferences being drawn from them when the sleeper wakes up. The art of divination, thus conceived, though it depends upon experiences during sleep, is practised in the waking state, not necessarily by the dreamer himself, but by an 'interpreter' of dreams who can explain their significance (cf. 464b5-16). The distinction between ‘during’ and ‘through’ sleep thus points to two different aspects of divination: (1) the having of precognitive dream experience;

462024

ON DIVINATION THROUGH

SLEEP

157

and (2) the interpreting of one's own or others' dreams, with a view to drawing inferences from them. Plato draws the distinction at Timaeus 71e-72b. Aristotle does not draw it explicitly, but it needs to be kept in mind. When he is discussing the supposed 'powers' of certain subjects, he is thinking of (1). When he discusses the significance of dreams for medical diagnosis, he is thinking of (2). The translation ‘divination’ is preferable to ‘prophecy’ (Hett), since the latter is limited to the foretelling of future events. While Aristotle's frequent references to 'prevision' (e.g. 462b25, 463b15) show that he usually has in mind foreseeing of the future, his discussion has some bearing upon paranormal cognition more generally. The ‘wave theory’ that he will advance later (464a5—b5) has, in fact, more relevance to telepathic cognition of past or present events than to foreseeing future ones. See note 51 on 463b31-464a19. b24-25. At the Pillars of Heracles: Gibraltar; On the Borysthenes: the Dnieper. Note the special difficulty of understanding cognition at remote distance. These places, at the extremities of the world known to the Greeks, would be paradigm locations for events occurring 'far away' (cf. 463b2-3). Such events would be ‘outlandish in place’ (464a2). Although Aristotle speaks of 'prevision' of future events, his reference to distant places points to the problem of explaining direct cognition of any event, past, present or future, that occurs beyond the ken of ordinary perception.

45. 462b26—463a3 Aristotle here distinguishes three possible ways of accounting for veridical dreams: they must be either (1) 'causes', (2) 'signs', or (3) 'coincidences'. What is the

basis

of this distinction,

and

what

ranges

of dream

are the

different

possibilities meant to cover? Through the example of the sun's being eclipsed while someone is taking a walk (b31—a2), Aristotle explicates 'coincidence' in terms of a pair of events, neither of which is a cause or a sign of the other.

Thus, if coincidence obtains

between A and B, then A cannot be either a cause or a sign of B, or vice versa. Conversely, if A is either a cause or a sign of B (or vice versa), then coincidence cannot obtain between A and B. For 'coincidence' see also Glossary, s.v. sumptoma.

It follows that for any pair of events, A and B, if A is either a cause or a sign of B or a coincidence, then A must be either a cause or a sign of B, or neither.

That, presumably, is why it is said to be 'necessary' that dreams must fall into one or more of these categories (b26). b28.

Three options are mentioned with respect to ‘causes’, 'signs', and 'coincidences': dreams must be 'either all or some of these, or one only'. Aristotle means that dreams,

taken

collectively,

may

fall into all three,

two,

or only

one

of his

158

NOTES

463a2

categories. Amongst these options, he will endorse the first: some dreams can be causes, some signs, while most are coincidences (463a30-b1). Huby

(55) takes ‘coincidences’ to cover ‘false dreams’, i.e. those that remain

unfulfilled, as well as those that are true only by chance. Admittedly, the notion of ‘coincidence’ as explained above would be wide enough to include dreams paired with any events whatever, provided they were neither 'causes' nor 'signs' of those events.

Thus, my dream last night of a blizzard in Moscow

would be

‘coincidence’ in relation to the Queen's having a boiled egg for breakfast this morning. But in any specific context the notion of coincidence is far narrower, since it is not any and every conjunction of items that will be of interest to us. In the case

of dreams,

the conjunctions

that interest us are, of course,

those

which hold between the dream and any real event alleged to be its ‘fulfilment’. Where there is no such event, then we do not have two items that could be thought of as ‘coinciding’. ‘False dreams’, then, will have nothing with which to coincide. If this is correct, Aristotle is not here thinking of dreams generally, in relation to innumerable concurrent or future events with which they are totally unconnected; he is thinking only of supposedly 'precognitive' dreams in relation to the events in which they are ostensibly fulfilled. Beare translates ‘the dreams’ at 026--27 as 'the dreams in question’. This goes beyond the text, but is a correct gloss upon it. In that case, the three types of explanation apply only to 'precognitive' dreams, with which the present treatise is mainly concerned. At the start of chapter 2, it is true, Aristotle will speak of dreams ‘in general’ (463b12), but even there his interest lies in accounting for veridical dreams that might lend some credibility to divination. a2-3.

Aristotle draws from his explication of ‘coincidence’ the corollary that 'no coincidence happens either always or for the most part’. This corollary follows, because it is universal or regular conjunctions of A-type events with B-type ones that warrant treating an A as a ‘cause’ or a 'sign' of a B, or vice versa.

Consequently, if 'coincidences' obtain only between events that are neither causes nor signs of one another, they can never instantiate an invariable or a regular conjunction. This point will be picked up later and applied to dreams that have no fulfilment (463b9-11, cf. b22-23). For the relation between ‘coincidence’ or 'chance' and the absence of regular conjunction, cf. Phys.

196b10-17, 198b32-199a3. Although Aristotle does not consider the options ‘all, some, or one only’ in relation to dreams taken individually, we may ask how any particular dream would fare in terms of them. Could a given dream be explained in more than one of his three ways? Or are ‘causes’, ‘signs’ and ‘coincidences’ mutually exclusive? For a given dream in relation to any single event, the first option is clearly ruled out. Since ‘coincidence’ excludes both 'cause' and ‘sign’, no dream could

46343

ON DIVINATION THROUGH be a cause, a sign, and a coincidence,

SLEEP

159

with respect to one and the same event.

For the same reason, the second option would be partly ruled out: no dream could be both a cause of an event and coincidental with it, or both a sign of an event and coincidental with it. But might a given dream be both a cause and a sign of the same event? Aristotle illustrates 'causes' and 'signs' of an eclipse of the sun and of a fever, with quite different items (b28-31): as the 'cause' of a fever he gives ‘fatigue’, whereas for a 'sign' he gives 'roughness of the tongue'; similarly, for the 'cause' of a solar eclipse, he gives 'the moon’, but for a 'sign' he gives ‘the star's [i.e. the sun'sj entry into shadow'. Still, he does not define 'cause' and 'sign' in mutually exclusive terms. We might therefore suppose that a single item could be both a cause and a sign of one and the same event, e.g. a cloud of a rainstorm. Nevertheless, the examples to be given below of dreams as 'causes' and as 'signs' would hardly square with the idea that a given dream might be both cause and sign of the same event. Dreams may be 'signs' of incipient ilinesses or other bodily states (463a4—21); but those states will not be caused by the dreams that point towards them. Dreams may also be 'causes' of subsequent actions that they suggest to the dreamer (463a21-30); but since one does not normally infer one's own future intentional actions from signs, such dreams could hardly be signs, at least for the dreamer, of the actions they cause.

It seems, then, that

Aristotle would classify any given dream, with respect to any given event, in only one of his three ways. When he says (431) that ‘it is possible for some of the dreams to be both signs and causes', he means only that some veridical dreams can be explained as 'signs', and some as 'causes'. He does not imply that any are explainable in both ways. This would not preclude the same dream from being classified differently in relation to different events. Thus, if someone were to dream of having a heartattack and visiting a doctor during a thunderstorm, the dream might, in Aristotle's terms, be a 'sign' of a real illness, a 'cause' of a real medical visit, and

ἃ 'coincidence' in relation to a thunderstorm that happened to break out at the time of the visit. But Aristotle nowhere considers a complex case of that sort. 46. 463a3-21 The view that dreams should be taken seriously as symptoms in medical diagnosis is Hippocratic. See Introduction, 30. Aristotle is thinking of physical rather than mental disorders. For the conditions of which dreams are said to be ‘signs’ are "illnesses and other affections imminent in our bodies' (463a18-20). Aristotle has no systematic theory linking specific dream contents with types of personality or emotional state. He once floats the idea that the dream appearances of good people are 'better' than those of ordinary ones (EN 1102b10-11), perhaps simply because dreams will often re-enact their subjects' waking behaviour (cf. 463a23—27).

Cf. Introduction, 12, n. 23.

160

NOTES

463a7

a7-21. Aristotle here offers a plausible explanation of why internal movements should be more noticeable during sleep than in waking hours. The point is illustrated by sleepers who think it is thundering and lightning, or that they are tasting honey, or walking through fire (a12-16). Aristotle does not call these phenomena 'dreams', nor could he do so consistently with his official theory that dreams are by-products of prior waking perception of the external world (DI chs. 2-3). But any satisfactory theory of dreams ought to take these common experiences into account. See note 42 on 462a15-31.

47. 463a21-31 Aristotle now shows how dreams may be ‘causes’. They may bring about their own fulfilment, by suggesting a course of action that the dreamer subsequently adopts. This causal relation is compared with the converse one holding between the subject's actual or intended waking actions and his subsequent dreams. The tracing of dreams to the subject's waking preoccupations was a commonplace of rationalistic accounts of dreaming in antiquity. It is vividly illustrated in Lucretius IV. 962-86. See also Introduction, 8-9. Such an explanation is not wholly compatible with Aristotle's own official theory (DI chs.2—3) that dreams are a by-product of waking perception. For if someone dreams of an action that he merely intends to perform, but has not yet carried out, such a dream could hardly be due to a residue of waking perception. There could be no such residue from perception of an event that has, ex hypothesi, not yet occurred. Aristotle

does

not, indeed,

invoke

his official

theory

to explain

how

such

dreams occur. He says, vaguely, that the movement 'has had the way paved for it by daytime origins’ (425-27), but he does not describe the mechanism for this 'paving of the way'. It seems that dreams of this type, although common, were not the ones that he had in mind, when framing his official theory. For that theory was dominated by a visual model, in which the dreamer was a passive spectator rather than an active participant. See Introduction, 6-8. a25.

Aristotle here speaks of our performing actions 'in direct dream-vision'. For this expression, see Glossary, s.v. euthuoneiria. Aristotle does not suppose that such a dream involves clairvoyant prevision of the subject's own future action. Nor need he concede to a believer in 'mantic' dreams that the subject's subsequent action 15 infallibly predictable from the dream's occurrence. For the dream in no way predetermines subsequent action, since any voluntary action 1t may suggest will (on Aristotle's view) still be up to the agent to perform or not to perform when he wakes up. If he dreams, say, of going to the doctor, it will remain up to him, when he wakes, to decide whether to go or not. His dream will be neither a necessary nor a sufficient condition of his going, if we suppose, as seems reasonable, (a) that he might have gone, even had he not

463a31

ON DIVINATION THROUGH SLEEP

161

dreamed of doing so; (b) that he might have dreamed of going, yet decided not

to go. A dream might, indeed, serve as a warning, causing the subject not to do something he would otherwise have done. Aristotle does not consider this possibility,

because

he is concerned

here

with

cases

in which

the dream

is

ostensibly fulfilled. Nevertheless, the possibility shows that no dream could be more than a contributory cause of the agent's acting in accordance with it.

Hence, if divination from dreams entails cast-iron prediction of the subject's future actions, no dream prediction.

could provide sufficient basis for making

such a

48. 463a31-b11 Dreams that prove to be veridical are here compared with chance mention of some actual event which happens to occur, or of some person who just happens to turn up. Mention of the person ‘is neither a cause nor a sign of his turning up' (b6—7), if (as we are to suppose) the speaker has no reason to believe that he will appear, and mention of him does not (like ‘talk of the devil’) cause him to

appear. The mention of him is, then, 'coincidence' in relation to his appearance. Similarly,

Aristotle suggests, most

veridical dreams

are mere

coincidence

in

relation to those events in which they are fulfilled. This undercuts the viewing of such dreams as precognitive, since they need no more contain genuine precognition than does a chance reference to someone who happens to turn up. It need not, therefore, be supposed that the dreamer has any foreknowledge or prevision of the event in question. Compare, also, the reference to gamblers who are occasionally lucky (463b18—22). Aristotle will later supplement the ‘coincidence’ theory with the suggestion that certain subjects are especially prone, by nature, to have veridical dream vision (463b16-19, cf. 464a22-b5).

The comparison of a veridical dream with chance mention of someone who turns up is apt in the following way. A mere mention of someone by name is not, in itself, a prediction that that person will appear. Since mere mention of him does not purport to assert that he will appear, it would not be proved correct by his appearing (nor incorrect by the contrary). Likewise, a dream about someone does not amount to a prediction that anything in particular will occur with respect to that individual. For dreams do not achieve reference to the real world, in the way that verbal predictions do. b1-3.

Aristotle assigns to coincidence all dreams that are 'outlandish or concern matters where no causal initiative lies in the dreamers themselves, e.g. a naval battle or things taking place far away’. Cf. 464a1-2, where some dream events are characterised as outlandish ‘either in time or place or magnitude’. Presumably, a dream of a naval battle would exemplify an occurrence where ‘no causal initiative lies in the dreamer’. For nothing he can do will help to bring about or prevent the battle, or affect its outcome.

162

NOTES

463b9

A naval battle is used as an example of an event whose occurrence or nonoccurrence is, according to a famous fatalist argument discussed by Aristotle in De Interpretatione ch. 9, necessitated by its having been antecedently true that it would or that it would not occur. The similarity of the example is striking. Perhaps, as Huby (60) speculates, the argument of the De Interpretatione was a response to the problem raised in Aristotle's mind by an actual and correct prediction of a sea-battle, based on a prophetic dream. Aristotle

occurrence occur.

might

say,

of

such

a dream,

that

it no

more

necessitated

the

of the battle than did the antecedent truth that the battle would

Nor was the dream true at any time before the battle took place.

Rather,

the dream should be said to have 'come true' in virtue of the battle's occurrence. It was neither a sign nor a cause of the battle, but was mere coincidence. For there could be no connection between it and a distant event totally beyond the dreamer's cognizance or control. To dismiss such a dream as ‘coincidence’ would be in line with Aristotle's rejection of the fatalistic argument of De Interpretatione ch. 9. Dreams of ‘outlandish’ events are not easily accommodated by Aristotle's official theory that dreams are due to a residue from waking perception. A dream of a distant, and still future, naval battle could not be due to a residue from waking perception of that event. It might be due to distortion of a residue from some other waking perceptual experience (cf. 461a15-25, 464b10-16). But it would be no less plausible to attribute the dream to some quite different cause, e.g. the subject's interest in a battle of which he had read. Aristotle's account of the genesis of dreams seems too narrow to allow for causes of that sort. b9-11. It is argued, finally, that it is just because most veridical dreams are coincidences that many dreams are not fulfilled. ‘For coincidences happen neither always nor for the most part. For ‘coincidence’ (sumptoma) cf. 463223, and see note 45 on 462b26-463a3.

Chapter Two 49. 463b12-22 Although Aristotle starts this chapter by speaking 'in general' (b12), he is still primarily concerned with veridical dreams that givean appearance of having been precognitive. For the rejection of divine provenance, cf. 462b20—22, 464a18-22, and see Introduction, 9, 13-14. The present argument appeals partly to the fact that animals dream. For animal dreams, cf. HA 536b27-537b12 (Appendix), and see Introduction, 35, 51-2. Aristotle assumes

convey

divine

that if dreams

messages,

they

occurred 'for that purpose' (013--14), i.e. to

would

be

sent

only

to those

capable

of

463017

ON DIVINATION THROUGH

SLEEP

163

understanding them. Animals, lacking rationality, would be incapable of this. A similar assumption underlies the ensuing argument (b15~19, cf. 464a19-24), that if dreams were sent by God, they would be experienced by the intelligent, and not, as is actually the case, by ordinary people. These remarks reflect Aristotle's "intellectualis bias. But behind them lies the reasonable notion that communication implies a wish on the sender's part to be understood, and thus a choice of recipients capable of getting the message. The point is well elaborated by Cicero, who argues that the enigmatic character of many dreams would make them utterly unsuited to convey messages from anyone. If they are, indeed, ‘messages’ from the gods, they evidence sheer madness on the part of their senders (De Divinatione II, 131-4). Aristotle's point, however, is not merely negative. He also argues constructively, that dreams, although not divine, are nevertheless 'daemonic' (b14—15). This is no mere parenthetical remark, but is the conclusion for which the ‘proof of b15—22 is offered. The fact that veridical dreams are prevalent among ordinary people, it is argued, sits better with a psychological than with a religious explanation for them. An account of them in terms of their subjects' 'nature' suggests that dreams are, like nature herself, 'daemonic, but not divine'. For the interpretation of this remark and its relation to Freud, see Introduction,

section 10. The point is that the dreams of certain subjects are so numerous and varied that there will be enough random correspondence between their contents and real events to warrant calling them 'daemonic'. For although they have a wholly natural explanation and there is nothing supernatural about their origin, nevertheless, like other events in nature, they display an uncanny resemblance to the products of conscious design. On this interpretation, the 'daemonic' is closely related to 'luck' (zuche) and ‘the fortuitous'(to automaton), and should be understood in the light of Aristotle's discussion of those concepts in Physics Il. 4-6. A connection between 'the daemonic' and 'the fortuitous' is also suggested by the terms of Aristotle's initial questions about divination at 453b22-24. See note 1 on 453b11-24 and Glossary, s.v. daimonios. b17.

For the 'atrabilious', see Glossary, s.v. melancholikos. See also 461a21-25, where such individuals are said to be prone to grotesque dreams, and 464432-b5, where their tendency to have veridical dreams is again attributed to luck.

b20-21. Aristotle illustrates his hypothesis by quoting the words of gamblers in a game of dice. Their slogan could be freely translated 'If you throw often enough your luck will change’ (thus Beare and Hett). If gamblers in Aristotle's day committed the so-called 'gambler's fallacy', this may have been whet they meant by it. But Aristotle is not endorsing or committing that fallacy himself. For he is not suggesting that a run of bad luck increases the chances of success on each subsequent throw. He is merely arguing that, given the number and

164

NOTES

463b22

range of dreams experienced by certain types of subject, it is not improbable that they should sometimes correspond with actual events. See Introduction, 43—4, with n. 88.

50. 463b22-31 Aristotle now says that it is not surprising that many dreams are not fulfilled. For the same holds true with respect to signs and causes in other spheres, e.g. with medical and weather warnings. Why does he make this point, when he has twice noted it as an implication of putting most veridical dreams down to coincidence? To view them as such is already to imply that they are exceptional (463b9—11, cf. 463a2—3), so it is not clear why this has to be pointed out again here. The passage fits better with the thought of the previous chapter (and so may be out of place). Aristotle is reverting to the two kinds of precognitive dream that he had previously allowed, i.e. those that are ‘signs’ or 'causes' (463a3-31). He now points out that signs or causes of imminent

events, e.g. of weather changes or human

actions, do not

always guarantee their occurrence. For countervailing factors may intervene to prevent what would otherwise have happened. For example, an impending storm may be averted by a change of wind, or an illness by medical treatment. Similarly, a dream may be a sign of an illness that was prevented, or a cause of an action which the subject was prevented from carrying out. Hence, to hold that dreams can be signs or causes is consistent with the frequent nonoccurrence of the events they presage or produce. Perhaps Aristotle is anticipating an objection that since dreams are frequently unfulfilled, they cannot be 'causes' or 'signs' at all, given the normal implications of those terms. To such an objection he replies, in effect, that A may be a sign or a cause of B, even when B fails to occur (b29-31). For A's being a sign or cause of B does not entail that B will inevitably occur, but only that it is 'going to' occur, i.e. that it will occur but for the intervention of some countervailing factor. For example, if a woman were to dream that she is pregnant, that she visits a doctor, and gives birth to a child, Aristotle might regard her dream as a 'sign' that she was going to give birth, and as a 'cause' of a medical visit, even if she were to meet with a fatal accident while on the way

to the doctor.

Moreover,

paradoxically, the same dream might be viewed both as a 'sign' of an impending event, i.e. the birth of the child, and as a 'cause' of its non-occurrence, e.g. if it

led her to seek an abortion.

But Aristotle does not observe that dreams may

cause actions which prevent their fulfilment.

He merely argues that non-

fulfilment need not preclude them from being causes or signs.

b25-26. For if another movement should take place, prevailing over the one from which (when it was going to happen) the sign occurred, then the latter The intervening 'movement' (for which see movement does not occur:

463b28

ON DIVINATION THROUGH

Glossary,

s.v. kinesis)

inoperative,

but

over

SLEEP

165

is said to prevail, not over the sign that it renders the

imminent

event

that it forestalls.

Note

also

that

impending, but not yet actual, events can give rise to 'signs': the sign is said to occur ‘from’ the prospective event, when the latter was still merely ‘going to’ happen. The clouds occur 'from' the impending storm. b28-29.

Not everything that was going to happen actually does happen: For this

distinction cf. GC

337b4—10,

and Williams,

197-8.

See also Glossary,

s.v.

mellein. What is 'going to' happen is what will happen unless something should intervene to prevent it. A pregnant woman is 'going to' give birth to a child, but it need not be the case that she "will' give birth, since she may suffer a miscarriage. A house on fire may be 'going to' burn down, but it need not be the case that it ‘will’ burn down, since a rainstorm may put the fire out. Aristotle's distinction, though difficult to make precise, embodies a common-sense view of events as both causally ordered and contingent. Of many events in the natural order it is not true to say in advance that they will inevitably occur. For it is often conceivable,

however

unlikely,

that some

new

factor will intervene

to

prevent them. The distinction between what ‘will’ happen and what is 'going to' happen (b29) has some bearing upon the fatalist argument discussed in De Interpretatione ch. 9. See note 48 on 463a31-b11. A dream could, with respect to a natural event, presage what was 'going to' happen, but there will often be room for such a dream to remain unfulfilled. For, as we noticed earlier (see note 47 on 463a2131), the fulfilment of a dream will often depend upon further factors, including (where the event presaged is a human action) the agent's own choice. Thus, Aristotle's distinction between what 'will’ happen and what is ‘going to' happen is in keeping with his refutation of fatalism in the De Interpretatione. It remains doubtful, however, whether the distinction does in fact help that refutation. Williams (loc. cit.) has argued that its seeming relevance to fatalism is illusory. b29-31. There are causes of a certain kind, from which no fulfilment ensued, and these things are natural signs of certain things that failed to occur: The reference of 'these things' is not clear, and it may be wrong to suggest, as the present translation does, that unfulfilled causes are themselves natural signs of non-occurring events. We might indeed suppose that one and the same item could be both a sign and a cause, of the same event, e.g. a cloud of a rainstorm. But Aristotle has not previously treated 'causes' and 'signs' as overlapping classes (see note 45 on 462b26--463a3), nor has he viewed the former as a subclass of the latter. Weather warnings and medical symptoms were given as examples of unfulfilled 'signs', and frustrated human plans as cases of unfulfilled 'causes'. Possibly, therefore, the two parts of the present sentence

166

NOTES

463b31

refer to 'causes' and 'signs' respectively. The reference of 'these things' will then be to the medical and weather signs mentioned at b23—25.

51. 463b31-464a19 Aristotle now supplements the preceding explanations of veridical dreams with an alternative theory, preferable (in his view) to the theory put forward by Democritus, one of the earliest proponents of Greek atomic theory, and the intellectual forerunner of Epicurus and Lucretius. Dreams were attributed by the atomists to 'images' (eidola) or 'films' (Latin simulacra), which are continually being given off by physical objects, and which pervade the environment in a huge variety of combinations. Through amalgamation and distortion, due to collisions of their component atoms, these ‘films' may assume

fantastic forms, while retaining certain physical and psychological features of the real figures from which they originated. Usually they go unnoticed, but when they penetrate the body during sleep, they may impinge upon the mind, itself a material entity composed of atoms, and thus cause dream experience. Democritus' version of this theory is alluded to by Sextus Empiricus (adv. Mathem. 9.19), and is further evidenced in two passages from Plutarch (Quaestiones Conviviales) 682f-683b, 734f-735c). But the theory is most fully elaborated by Lucretius (IV. 722-822, 962-1036). See also Glossary, s.v. eidolon. It is notable that an atomist should be the only previous theorist whom Aristotle mentions by name.

His own account of dreams has, in fact, much in

common with the atomists'. Like them, he sought a naturalistic explanation of dreams by tracing them to waking experience or behaviour. He also dismissed

belief in their divine provenance:

even if dream phenomena should causally

explain popular belief in anthropomorphic gods, they could not rationally justify it Cf. Introduction, 13-14. Nevertheless, he firmly rejected the purely materialist conception of the mind, and the wider account of perception and imagination, upon which the atomic theory of dreams was grounded. b31-a2.

As for dreams that do not contain causes of the kind we have described,

but outlandish ones, either in time or place or magnitude:

(b31—a2).

Ross

(281) glosses this: 'with regard to dreams that have not such causes, but causes extraordinary in respect of time or place or magnitude’. This wrongly suggests that Aristotle is talking about dreams that ‘have extraordinary causes’. But he is

clearly referring to dreams with a certain type of content.

'Causes of the kind

we have described’ are not causes of dreams, but dream content having causal impact upon subsequent events, by affecting the dreamer's own waking actions. Such dream content is in no way 'outlandish'. It includes such actions as the subject might himself perform, so that the impact of such dreams upon real events is intelligible. By contrast, a causal connection is not intelligible, where dreams are 'outlandish in time or place or magnitude'

463031

ON DIVINATION THROUGH

SLEEP

167

By ‘outlandish in time or place' Aristotle means that the temporal or spatial location of the dream event is so remote from the dreamer's own as to rule out any possible connection between the dream and its fulfilment. If someone in Athens were to dream of a sea-battle occurring a century later at the Pillars of Heracles, his dream would contain ‘causes outlandish in time and place’.

For no

one would suppose that such an event was predictable from facts already known to the dreamer or could occur through his agency. He could neither infer nor bring about an occurrence at such a remote time and distance, and so could have

had no genuine foreknowledge of it. It is less obvious what is meant by ‘outlandish in magnitude’. But perhaps Aristotle is thinking of cataclysmic events such as earthquakes, that are beyond all human power either to produce or to prevent. This would naturally lead him to add ‘or in none of those ways, but where those experiencing the dreams still have no causal initiatives within themselves’. For this addition would cover cases where the event dreamt of is neither remote in time or place, nor totally beyond human control, but where no action within the dreamer's own power could bring about or prevent its occurrence, e.g. a particular athlete's winning a contest (cf. EN 1111b24). In such cases, Aristotle thinks, 'prevision' may still be ascribed to coincidence

(a4—5). But he also suggests an alternative explanation, postulating wave-borne impulses, akin to disturbances propagated in water or air. Here he uses the same mechanical model for the transmission of change as he had used for his own general theory of dreams. Compare 464a6-9 with 459328—b5. But there is a significant difference between the two uses of the model. In the general theory of dreams, Aristotle could plausibly use it to show how traces of physical changes initiated in the subject's waking experience might persist within him during sleep. In the present case, however, the model is supposed to explain the dreamer's awareness of events occurring at remote distance in space and time. For it purports to account for 'previsions of what is going to happen, even on matters

of the sort mentioned

above'

(al8-19),

i.e. the ‘outlandish’ cases and

others mentioned at al-4. The theory is hard to understand in connection with occurrences in the remote future. For how is it supposed to account for prevision of events that have not yet occurred? If the impulses proceed, as Aristotle says (all), from the objects "whence Democritus derives his "images" and "emanations" ', i.e. from material objects in their present or previous state, how do they enable the dreamer to foresee the remote future? The theory is, in fact, better suited to telepathy than to precognition, as is the hypothesis of special rapport between distant friends (464a27-32). Aristotle's initial question as to whether and how certain people could genuinely foresee the future (462b24—26, cf. 453b20—22) remains virtually untouched. Cf. Huby, 60.

168

NOTES

464a19

52. 464a19-27 Whatever the defects of the 'wave' theory (and it is too briefly sketched for us to understand its workings in detail), it does accommodate two features of dreaming which religious accounts leave unexplained: (1) it explains why dreams occur at night rather than by day; for the hypothesis that stimuli are more noticeable at night, cf. 464a12-17 with 460b27-461a8, 463a7-21; (2) it explains why such dreams occur to ordinary people rather than to the most

intelligent. Cf. 464a18-24 with 462b21—22, 463b15-18. It remains unclear, bowever, just how the wave theory is related to the general theory of dreams given in the De Insomniis, and indeed whether it is consistent

with that theory. If precognitive or telepathic dreams occur as a result of waveborne impulses reaching the subject during sleep, can Aristotle consistently maintain his formal definition of a dream as 'an appearance that arises from the movement of the sense-impressions' (462a29-30)? For that definition ascribes dreams to an internal residue from previous waking perception. The wave theory suggests origins of quite a different sort, viz concurrent impulses from external sources impinging upon the sleeper's consciousness. 53. 464a27—b5 Aristotle has now to explain, given the wave theory just advanced, (a) why a given item emitting 'waves' should be picked up only by some recipients and not others; and (b) why a given recipient should pick up signals from some items but not from others. Precognition is not available to everyone, but only to certain people, and it concerns not any and every item emitting waves, but only those of special interest to the dreaming subject. The theory of mutual rapport between friends can accommodate both these points. But could it have wider application? How far could precognitive dreams be explained in terms of the subject's greater familiarity with the items dreamt of? In effect, Aristotle postulates 'a selective factor at work' (Dodds, 120). Just as friends seen at a distance are most quickly recognized by each other, because of their mutual familiarity, so their 'movements' (431), i.e. the 'waves' propelled from

them,

are more

familiar

to those

who

know

them,

and

are thus

more

readily picked up. This suggestion might explain why a precognitive dream is more likely to concern a distant friend than a stranger close at hand. But it leaves unexplained how the sleeper gains any awareness of what is happening to a distant friend, or how any prevision of the friend's future could be gained from waves emanating from him. a27.

People who have direct dream-vision: See Glossary, s.v. euthuoneiria. Aristotle's remarks about the intuitions of the 'atrabilious' (a32-b1) address the same problem as his hypothesis about rapport between friends. They explain why precognition is limited to certain people. For 'atrabilious', cf. 461a22, 463b17,

and see Glossary, s.v. melancholikos.

Aristotle draws

on his earlier

suggestion (463b18—22) that such subjects strike lucky more often than most,

464427

ON DIVINATION THROUGH

SLEEP

169

because of the abnormal frequency of their visions. Owing to their ‘mutability’, ie. the rapidity with which their thoughts change (a33-b1), they envisage so many things that some of these are likely to prefigure actual events. They are thus 'good guessers' (433), compared with marksmen who shoot from a . distance, unwilling ‘to wait till the object of their speculation is close at hand’ (Beare), Thus their premonitions will occasionally prove correct. For 'good guessing’, cf. EN 1142b3, where it said to involve no reasoning and to be done quickly. In this way the 'intensity' of the atrabilious can explain their success in guessing. It is less easy to see how it explains why 'one movement does not get knocked out of them by another' (b4—5). For, given the rapid succession of their thoughts, it might be expected to have just the opposite effect. Moreover, it was suggested above (a24—27) that in some forms of insanity, subjects are especially susceptible to alien movements, precisely because their own movements 'get beaten off.

If so, how can the fact that one movement does not get knocked out

of them by another now help to explain successful precognition? Perhaps Aristotle means that, in their case, stimuli coming from external sources do not get repulsed by those from within their own bodies. Because these subjects concentrate intensely upon the external stimuli, these are not displaced by others, but have a specially strong impact upon their awareness. There remain the difficult lines (b1-4), in which the rapid sequence of thoughts in the atrabilious is likened to the ramblings of madmen. The text of these lines is conjectural (see textual note 25), but if correct, it refers to a fourth-

century poetess, Philaenis, who is cited in later Greek literature as the author of a notorious pornographic sex manual. See Athenaeus Deipnosophistae 220f, 335d-e,

457e,

Lucian

Pseudologista

25,

[Lucian]

Amores

28,

and

Polybius

12.13 with commentary by F.W. Walbank (Oxford 1967) vol. II, ad loc. Despite Ross (283), no compositions by Philaenis or fragments of her work are extant. Ross cites Athenaeus 335c-d. However, the verses there quoted are not attributed to Philaenis,

but to a different poet, Aeschrion,

who

protests,

assuming her voice, that the foul works imputed to her were written by someone else. The same verses also appear in the Greek Anthology (VII. 345), as does another poem protesting her innocence (VII. 450). But if Aristotle did mean to compare the words of madmen with her supposed writings, the meaningless words 'Aphrodite, -phrodite' can be explained as part of a nonsense jingle based

purely on assonance. B.A. van Groningen (Mnemosune ser 4.1 (1948) 107—8) has adduced papyrus evidence of an amulet or charm dating from the sixth century A.D., which includes the name 'Aphrodite', repeated eight times in successively shortened forms, one letter being dropped from the beginning of the word with each repetition:

'Aphroditen, -phroditen, -roditen, -oditen, etc’.

170

NOTES

464b5

Aristotle does not say that these utterances occur in any work by Philaenis, even though her alleged sexual themes have an obvious connection with Aphrodite, the goddess of love. All that matters in the example is the assonance of the successive word-endings. Just as such words are linked by mere sound rather than by sense, so the thoughts of the atrabilious are strung together without logical connection. For such an interpretation, see Ross 283-4. If that is correct, Aristotle is suggesting that mere random associations of ideas may give the impression that the speaker has a premonition of the truth. The point of comparing the atrabilious with madmen will be that in neither case does a mere sequence of words evidence any grasp of their meaning, let alone any basis for holding the beliefs they express. The words uttered may happen to be true, and relevant to some real situation, but the speakers cannot be said to

know what they are saying. Cf. EN 1147a18-22, b11—12. Thus, the precognitive dreams of the atrabilious are like the ramblings of madmen, who occasionally hit upon the truth without knowing it. This kind of scepticism with regard to divination can be traced back to Xenophanes (DK 21 fr. B34, as interpreted by Lesher, 6-17, 20, nn. 32-33, 21, nn. 36-38). Cf. also Plato's ironical comparison of politicians with diviners (Meno 99a-d).

54. 464b5-16 The 'premonitions' of subjects discussed in the previous paragraph have no rational basis, so divination from their dreams could not be a genuine skill. By contrast, Aristotle now says, the 'most skilled' interpreter of dreams is one who can ‘observe resemblances' (b6-7). Such an interpreter would be practising a skill, not merely riding hunches or making lucky guesses. But is this skilled interpreter one who can give the literal meaning of dream symbolism? Or does he merely discover the causal explanation of dreams, by tracing them to their origins in waking experience? Aristotle's remark that ‘direct dream-vision' can be interpreted by anyone (b7) suggests that he is thinking of causal explanation. A dream of Coriscus can readily be understood as due to a previous or prospective meeting with him in real life.

The causal link between the dream

such a case, be obvious.

and the event it rehearses will, in

Cf. ‘direct dream-vision'

of our past, present, or

intended actions (463a23—27), and see Glossary, s.v. euthuoneiria. By contrast, skill is needed

when

the dream is not a faithful reflection of its

original but contains distortion. In that case, the interpreter 'resemblances' between the dream and waking experiences, the dream's origin. To 'interpret' the dream is, on this view, what object or event experienced in waking life has Introduction, 41-2. Aristotle explains ‘resemblances’ by comparing dreams relation to the things that cast them. Having a dream of a

will have to discern in order to identify simply to discover produced it. See with reflections in man or a horse (cf.

464016 458b11) is the dream reflection. rather than

ON DIVINATION THROUGH

SLEEP

171

like seeing a reflection of a man or a horse. To identify the source of may be as difficult as identifying the source of a badly distorted Aristotle, unlike Freud, thinks of the distortion as having a physical a psychological cause (cf. 461414--25). For the limitations of

reflections as models for the understanding of dreams, see Introduction,

15-16,

55-6.

55. 464b16-18 This retrospect ties the three essays on sleep and dreams together, by claiming that the program outlined at 453b11—24 has now been completed.

173

Appendix Aristotle's Historia Animalium, Book IV, ch. 10! The Historia Animalium is a vast, and seemingly amorphous, collection of zoological information gathered by Aristotle and his associates from empirical observation. Historia in its title means ‘research’ or ‘inquiry’. The work records facts about animals which are fundamental for the projects of classification and explanation undertaken in Aristotle's other biological writings. Its date of composition is uncertain, but much of the research underlying it is now thought to have been carried out during Aristotle's middle years, before he returned to Athens in

335 B.C. and formed his circle at the Lyceum. The following chapter is of interest, both because it is specially relevant to the three essays in this volume, and because it admirably illustrates the painstaking, detailed observation of many species by which Aristotle's philosophical writings are informed. The reader may find it useful to refer to it in connection with section 8 of the Introduction, especially pp. 35-7, which contain some general remarks upon its content. Except

where noted,

the Greek

text

is that of A.L.

Peck

in the

Loeb Classical Library. The marginal line numbers are those of Peck's edition. The points of transition to a new Peck line are marked, every five lines, by vertical bars in the text and translation.

l

This Appendix is reprinted by kind permission of the publishers and the Loeb Classical Library from ARISTOTLE Historia Animalium, Book IV (volume IX) translated by A.L. Peck, Cambridge, Mass., 1965.

174

536b

TON

ΠΕΡῚ

TA

ΖΩ͂ΙΑ

ΙΣΤΟΡΙΩΝ,

IV. 10

10. Περὶ δ᾽ ὕπνου καὶ ἐγρηγόρσεως τῶν ζώων, | ὅτι μὲν ὅσα πεζὰ καὶ ἔναιμα πάντα καθεύδει καὶ ἐγρήγορεν, φανερὸν ποιοῦσι κατὰ τὴν αἴσθησιν. πάντα γὰρ ὅσα ἔχει βλεφαρίδας, μύοντα ποιεῖται τὸν ὕπνον. ἔτι δ᾽ ἐνυπνιάζειν φαίνονται οὐ μόνον

ἄνθρωποι,

ἀλλὰ

καὶ

ἵπποι

καὶ

κύνες

καὶ

βόες,

ἔτι

δὲ

πρόβατα καὶ αἶγες καὶ πᾶν γένος: δηλοῦσι δ᾽ οἱ κύνες τοῦτο μὲν ἄδηλον, ὅτι δὲ τὰ ἔνυδρα, oiov οἵ τ᾽ μαλακόστρακα, κάραβοί τε ἐστι πάντα ταῦτα, φαίνεται

τὸ τῶν ζῳοτόκων καὶ τετραπόδων | τῷ ὑλαγμῷ. περὶ δὲ τῶν φοτοκούντων καθεύδουσι, φανερόν. ὁμοίως δὲ καὶ ἰχθύες | καὶ τὰ μαλάκια καὶ τὰ καὶ τὰ τοιαῦτα. βραχύυπνα μὲν οὖν δὲ καθεύδοντα. σημεῖον δὲ κατὰ μὲν

τὰ

^

^

M

Y

T

ὄμματα

x

ταῖς

t

οὐκ

bd

ἀλλὰ

A^

ἔστι

+

| γὰρ

#

καλουμένους

νῦν

δ᾽,

ἂν

#4

x

x

(οὐδὲν +

οἱ

M

γὰρ

Ν.

ἰχθύες,

^

Ww

Ψύλλους,

΄

ἔχει

x

πολλοὶ

x

εἰ t

κἂν

χρονίζωσιν,



προσπίπτοντες,

x

βλέφαρα

w

x

#

ὥστε

οὗτοι

^

^

τὸ

πλῆθος

πολλάκις ΄

οἱ



ἁλιεῖς €

μὴ ^

τῇ

διὰ

τοὺς ^

>

πολλάκις t^

^

καὶ

νυκτὸς z

γίνονται

καὶ

t

λαμβάνειν

+

ῥᾳδίως"

κατεσθίουσι 5

δ᾽

à

A

ἐν

τῷ

βυθῷ

^

γὰρ a

τῇ

χειρὶ

ἔστιν x

ἐκ

^

τῶν

ἐπιπεσόντα ^

λαβεῖν

ΚΑῚ



τὸ

iA

δέλεαρ

x

ὥσπερ

,

τοιῶνδε τοῖς

πατάξαντα

^

μᾶλλον

ἰχθύσι ^

Aadeirv'

λαθεῖν τ

|

Y



Y

ἔτι

8’

1

3

ἐν

^

ταῖς

,

πυρίαις

CN

ἁλίσκονται

x

διὰ

ὑπὸ

*

τὸ

10

σφαῖραν

«

Ν

ἔστι οὕτως x

x

δὲ

τὸν

καιρὸν τοῦτον ἠρεμοῦσι σφόδρα, καὶ κινοῦσιν οὐδὲν πλὴν ἠρέμα τὸ οὐραῖον. δῆλον δὲ γίνεται ὅτι καθεύδουσι καὶ ταῖς φοραῖς, ἄν τι κινηθῇ ἡσυχαζόντων αὐτῶν: φέρονται γὰρ ὥσπερ ἐξ ὕπνου ὄντες.

537a

^

φθεῖρας Á,

χειρὶ

τῆς Y

OVTES.

περὶ

~

συνεχομένων αὐτῶν. ἀλλ᾽ τεκμήρασθαι ὅτι καθεύδουσιν] ὥστε

30

αὐτῶν),

^

θαλάττης, καὶ τοσοῦτοι τὸ πλῆθος ὥστε καὶ τὸ δέλεαρ, ὅ τι ἰχθύος ἦ, ἐὰν χρονίσῃ ἐπὶ τῆς γῆς, κατεσθίουσιν: καὶ

ἀνέλκουσι kj

΄

ἀτρεμίαις.

TOUS

τῆς àv

3

λαβεῖν

Y

[ἁλίσκονται ^

4

^

25

n

καθεύδειν.

Cf. DD 463bl2.

2.

See Introduction, pp. 51-2.

3.

The text of the bracketed passage is badly confused, and some editors would excise the whole paragraph. It fits poorly into the context, and is largely irrelevant to the argument, which flows smoothly if it is omitted. The textual problems are discussed in detail by Peck, 362-7.

15

536b

175 HISTORY OF

25

ANIMALS IV. 10.

As for sleep and waking of animals, | all creatures that have legs and are sanguineous give observable evidence that they sleep and wake. Thus, those that have eyelids go to sleep with them closed. It appears, moreover, that it is not only human beings who dream,!

but also horses,

30

dogs and cows, as well as sheep and goats, and the whole | family of viviparous quadrupeds. Dogs show it by whining in their sleep.? As for oviparous creatures, though it is not certain that they dream, it is 537a obvious that they sleep. | The same holds also for water-creatures, such as fish and

5

10

15

molluscs

and

crustaceans, crayfish

and

the like.

All those

creatures sleep only for short spells, no doubt, but it is evident that they do sleep. One cannot prove the point by reference to their eyes, since none of them has any eyelids, but only from periods when they are motionless. [For fish are caught, | unless because of sea-lice and so-called ‘fleas’, so that one can easily grasp them in the hand. But in fact, if they remain motionless for a long time, those creatures fall upon them at night in huge numbers and devour them. They are formed in the depths of the sea, and are so numerous that they devour even the bait if itl consists of fish, and remains for long on the sea-bed; and fishermen often haul up a sort of ball of them clinging to the bait. But it is rather from the following indications that one may infer that they sleep.] Thus it is often possible to attack fishes unawares, so that one can catch them in the hand, or to deal them a blow unawares. | During these times they keep perfectly still and do not move at all, except for a slight motion of the tail. That they are asleep also becomes apparent from their movements, if any disturbance occurs while they are resting. For then they start up as if roused from sleep. Moreover, the fact that they can be captured by torchlight is due to their being asleep. Often,

176

TON

ΠΕΡῚ

TA

ZOIA

πολλάκις δὲ καὶ oi Ouvvookómot δῆλον δ᾽ ἐκ τοῦ ἡσυχάζοντας ἁλίσκεσθαι. {

LA

74

οὕτως

t^

ὥστε

καθεύδουσι

δὲ

[4

x

Pd

βαλλόντων

μὴ

b

τῆς

καὶ

IXTOPION

περιβάλλονται καθεύδοντας: τὰ λευκὰ ὑποφαίνοντας

νυκτὸς

^

537a

μᾶλλον

>

^

κινεῖσθαι.



-

bj

*

τὰ

δὲ

τῆς

Ww

^

^

πλεῖστα

ἡμέρας τ

,

#

καθεύδουσι

τῆς γῆς ἢ τῆς ἄμμου ἢ λίθου τινὸς ἐχόμενοι ἐν τῷ βυθῷ, ἢ ἀποκρύψαντες ὑπὸ πέτραν ἢ | θῖνα ἑαυτούς, οἱ δὲ πλατεῖς ἐν τῇ ἄμμῳ: γινώσκονται δὲ TH σχηματίσει τῆς ἄμμου, καὶ λαμβάνονται τυπτόμενοι τοῖς τριώδουσιν. λαμβάνονται δὲ καὶ λάβραξ καὶ χρύσοφρυς καὶ κεστρεὺς καὶ ὅσοι τοιοῦτοι τριώδοντι t

LA

ἡμέρας

΄»

πολλάκις

x

διὰ

^

τὸ

£^

καθεύδειν:

+

εἰ

*

δὲ

,

μή,

,

^

οὐδὲν

δοκεῖ

^

A

καθεύδει δὲ καὶ τὰ μαλάκια τὸν αὐτὸν τρόπον | ὅνπερ οἱ ἰχθύες: ὁμοίως δὲ καὶ τὰ μαλακόστρακα τούτοις. καὶ τὰ ἔντομα δὲ τῶν ζῴων ὅτι τυγχάνει ὕπνου, διὰ τοιούτων σημείων ἐστὶ φανερόν: ἡσυχάζουσί τε γὰρ καὶ ἀκινητίζουσιν ἐπιδήλως. μάλιστα δ᾽ ἐπὶ τῶν μελιττῶν τοῦτο δῆλον: ἠρεμοῦσι γὰρ καὶ παύονται βομβοῦσαι τῆς νυκτός. δῆλον δὲ καὶ ἐπὶ τῶν ἐν ποσὶ

The tunny-fishing industry seems to have been well organized. 'A man was posted on a high place, from which he could see the shoals coming, and so make a sign to the fishermen to let down their nets, — like the hooer in the Cornish pilchard-fishery' (Liddell & Scott-Jones, Greek English Lexicon, 8th. ed. Oxford 1901, s.v. 8uvvookórtog).

5.

Cartilaginous creatures, such as sharks and dog-fish.

25

τῶν

τοιούτων ληφθῆναι àv τριίζώδοντι. τὰ δὲ σελάχη οὕτω καθεύδει ἐνίοτε ὥστε καὶ λαμβάνεσθαι τῇ χειρί. δελφὶς δὲ καὶ φάλαινα, [ καὶ ὅσα αὐλὸν ἔχει, ὑπερέχοντα τὸν αὐλὸν καθεύδει τῆς θαλάττης, δι᾽ οὗ ἀναπνέουσιν ἠρέμα κινοῦντες τὰς πτέρυγας" καὶ δελφῖνός ye καὶ ῥέγχοντος ἤδη ἠκρόανταί τινες.

4.

20

30 537b

10

5378

HISTORY OF ANIMALS, IV. 10

177

too, the tunny-watchers^ | get the nets around them while they sleep: that is plain from their remaining still, and revealing their white underbellies, during capture. They sleep more at night than by day, and so soundly that they are not disturbed when the nets are being cast around them. In most cases fish sleep clinging fast to the sea-bed, either to the sand or to a stone on the bottom, or after hiding 25 themselves away under a rock or | mud-bank. Flat fish sleep in the sand; they can be recognized from their outline in the sand, and they are captured by being struck with a three-pronged spear. The bass, the gilthead, the grey mullet, and all such are often caught in the daytime with a spear, thanks to their being asleep. Otherwise, it would not seem 30 possible for fish of that sort to be caught with a | spear. Selachians? sometimes sleep so soundly that they can actually be taken by hand. 5370 The dolphin, the whale, and | all creatures that possess a blow-hole, sleep with the blow-hole protruding above the surface of the sea, and breathe through it, while gently moving their fins; some observers, in fact, have actually heard the dolphin snoring. Molluscs sleep in the same manner as | fish, and so do crustaceans. Similar indications prove that insects also are creatures that experience sleep. For they too can be observed at rest and motionless. This is especially clear in bees. For at night they keep still and cease their 10 buzzing. But it is also obvious in the | commonest varieties of insect. Not only do they rest at night because they cannot see clearly (since all 20

178

ΤΩΝ

ΠΕΡῚ

TA

ΖΩ͂ΙΑ

IZTOPION

537b

μάλιστα τῶν τοιούτων: οὐ γὰρ μόνον διὰ τὸ μὴ ὀξὺ βλέπειν ἡσυχάζουσι τῆς νυκτός (ἅπαντα γὰρ ἀμυδρῶς βλέπει τὰ σκληρόφθαλμα) ἀλλὰ καὶ πρὸς τὸ φῶς τὸ τῶν λύχνων ἡσυχάζοντα φαίνονται οὐδὲν ἧττον. ἐνυπνιάζει δὲ τῶν ζῴων μάλιστα ἄνθρωπος. καὶ νέοις μὲν οὖσι καὶ παιδίοις ἔτι πάμπαν οὐ γίγνεται | ἐνύπνιον, ἀλλ᾽ ἄρχεται τοῖς πλείστοις περὶ τέτταρα ἔτη ἢ πέντε: ἤδη δὲ γεγόνασι καὶ ἄνδρες καὶ γυναῖκες οἱ ὅλως οὐδὲν πώποτε ἐνύπνιον εἶδον. συνέβη δέ τισι τῶν τοιούτων προϊούσης τῆς ἡλικίας ἰδεῖν ἐνύπνιον, καὶ μετὰ ταῦτα γενέσθαι περὶ τὸ σῶμα | μεταβολὴν τοῖς μὲν εἰς θάνατον τοῖς δ᾽ εἰς ἀρρωστίαν. ἔχει

περὶ τὸν

μὲν οὖν τρόπον.

αἰσθήσεως

καὶ

ὕπνου

καὶ

ἐγρηγόρσεως

τοῦτον

6.

βλέπει, the reading of most mss, has been preferred to βλέπουσι printed by Peck.

15

20

HISTORY OF ANIMALS, IV. 10

537b

hard-eyed

15

creatures

have poor

179

vision)$ but even when

exposed

to

lamp-light they can be seen to go on sleeping just as soundly. Of all creatures human beings are the most given to dreaming. In young children and infants | dreams do not occur at all, but they begin in most cases around the age of four or five.

There

have, in fact, been

cases of adult men and women who have never once experienced a dream. It has also happened that certain subjects of that sort have

experienced 20

physical

a dream

| change,

late in life, and

leading

in some

have

thereafter

cases to death,

and

undergone in others

debility. So much then for perception, and for sleeping and waking.

to

180

Glossary account, see logos. activity, actualisation, see energeia.

affection, see pathos. aisthanesthai, ‘to perceive’, ‘to exercise perception’. This verb has a range of uses corresponding to those of the noun aisthésis (q.v.). It is used primarily of the exercise of the senses, or of the ‘common sense’ faculty, in ordinary waking experience (e.g. at 459b4, b8), and includes awareness of internal as well as external stimuli (454a3-4).

In an extended and qualified sense, it covers dreams. Dreaming subjects are treated ‘as if perceiving (461b27-28). Their experience is akin to perception, yet not due to normal sensory functioning. aisthéma, 'sense-impression'. A technical term, sometimes used like 'sense-datum' in modern philosophy, to mean a sensory impression present to a subject in any perceptual experience. aisthemata are also conceived as persisting in the senseorgans, or as leaving a residue therein, after the external stimulus has ceased to affect them (460b2-3). See also 456226, 461a19, 426, 461b22, 462230. aisthesis, 'sense', 'perception', 'sense-faculty'. This noun can mean

any or all of the

five senses, and in that use has been translated 'sense'. It is also used in connection

with the ‘common sense' (for which see DA 425a27, PA 686a32, DM 450a10-13a), the primary sense-faculty, of which the five senses are treated by Aristotle as separate modalities (cf. 455a20-22). Sometimes it means what occurs in the subject while the senses are being exercised, and in that use is translated ‘perception’. It can also bear a more general sense of 'awareness' or 'feeling', without reference to the exercise of any particular sense, as at 464a15. It thus comes closer than any other Greek word to the modern concept of 'consciousness'. aisthétérion, 'sense-organ'. A generic term for the bodily organs in whose workings the functioning of the senses consists. It

can also be used of the heart, sometimes

referred to as ‘the primary sense-organ' (e.g. 455a19-21), and regarded as a ‘master organ’, whose operation is necessary for the working of all the others (455a33-b2,

455b10-13). aisthétikon, ‘perceptual’, sometimes used with the article to, to mean 'the perceptual part’, though a word for 'part' is not always added. The phrase stands for what might be loosely called 'the perceptual apparatus’ or ‘the sensorium', without differentiating its psychological aspects from their physical basis. When Aristotle means the latter, he tends to speak of the 'sense-organ' (aisthétérion, q.v.).

GLOSSARY

181

aisthéton, 'sense-object'. Used of sensa, or sensible properties (colours, sounds etc),

that cause perception by stimulating the relevant sense-organs, but are themselves independent of the perceiving subject. See, e.g., 459a24-25, 460b2, 460b23.. aitia,

noun,

'cause',

'explanation',

'explanatory

factor',

reason';

aitios, adjective,

'causally responsible', 'explanatory'. These words have usually been translated either ‘cause’ or ‘explanation’. 'Explanatory factor’ is required at 455b14-16, where four types of factor, traditionally known as the four 'causes', are reviewed. For this key Aristotelian concept,

see An Post.

94320-96219,

Phys.

194b16-195a3,

198a14-21,

Metaph. 983a24-33, 1013a24-b3. The English ‘cause’ often suits the kind of aitia that Aristotle calls the 'source of change’ (traditionally, the ‘efficient cause’). ‘Cause’ is therefore the best translation when an explanation of that type is relevant, e.g. at 453b14, b17, b20, 455b9-10, 457b2, 462b27, 463a22.

See also arché, and note 8 on

455b13-34. alloiosis, ‘alteration’. That species of ‘movement’ (kinésis, q.v.), in which something undergoes qualitative change, e.g. of temperature. It is applied at 459b1-5 to the modification of the sense-organs which occurs during and after perception, and is said to be 'a sort of alteration'. alteration, see previous entry. apparition, see eidölon, phantasma. appear, appearance, see phainesthai, phantasia, phantazein. arche, ‘beginning’, ‘starting-point’, ‘origin’, ‘source’, ‘cause’. For Aristotle's account of the many connected uses of this word, see Metaph. V.1. In the present works, the most important of these is its use in conjunction with kinésis (q.v.) to mean ‘source of change’. In Aristotle's physiology, the. heart is often thought of in this way, being regarded as the originating point of vital processes, or as the ‘headquarters’ where perceptual stimuli are registered. Cf. PA 670226, where it is appropriately called ‘the acropolis of the body, as it were'. Elsewhere arché is used in a wider sense, to mean

an initiative that gives rise to change. In this use it comes close to ‘cause’, and has been so translated at 463b28, 029, 46441. atrabilious, see melancholikos.

automaton, ‘fortuitous’. Used at 453b24 of events due to chance (apo tautomatou). Aristotle's fullest account of this notion occurs in Phys. IL 6, following his analysis of tuché (luck). Both concepts characterize events leading to consequences such as might have been intentionally produced but were not in fact intended. "Τῆς fortuitous’ (to automaton) is the wider notion, since it is not limited (as is tuché) to rational human actions that have (from the agent's perspective) fortunate but unintended consequences; it includes the movements of non-rational agents or inanimate things, such as the bolting of a horse or the falling of a rock. Such events are said to be

182

GLOSSARY

fortuitous, if they have good or bad consequences that might have been intentionally produced, but were not thus produced (e.g. if a falling rock happens to kill someone). See Introduction, 44-5, and note 1 on 453b11—24.

become, see gignesthai. beginning, see arché. capacity, see dunamis. cause, see aitia, arche. change, see alloiösis, kinein, kinesis.

clairvoyance, see euthuoneiria. coincidence, see sumptoma. come to be, see gignesthai. consciousness, see aisthanesthai, aisthesis, aisthetikon.

consequence, see sumbainein. daemonic, see next entry. daimonion, 'daemonic'. This term is applied to dreams and to nature, which is said to be 'daemonic but not divine' (463b14). to daimonion, ‘the daemonic', is used at 453b23 for a factor that explains natural or fortuitous events, as distinct from those

due to human agency. In neither context does the word carry any connotation of the ‘supernatural’. Nor does 'mysterious' (Barnes) quite capture the meaning. More likely, the word is used here to characterize events that may look like, but are not in fact, the

products of conscious design. The English 'uncanny' contains a somewhat similar idea. See also note 49 on 463b12—22 and Introduction, section 10. definition, see logos. divination, see mantiké.

dokein, ‘to judge’, ‘to think’, ‘to seem’. It is often hard to decide how much intellectual freight this verb carries. In some places, especially 461b5-6, 462a1—2, it is sharply distinguished from phainesthai (q.v.), as signifying not merely that something appears to the subject to be the case, but that it is thought or believed to be so. In such places it has been rendered ‘judge’, in keeping with ‘judgment’, used for its cognate noun doxa (q.v.). But this contrast is not consistently observed in Aristotle's usage (compare 458b29 with 460b18), and where it is not in evidence, the weaker ‘seems’ has occasionally been preferred (e.g. 456b16, 460b26, 462b26). doxa, ‘judgment’. This word is often given as 'belief or ‘opinion’, but has here usually been translated ‘judgment’, beause of its kinship with dokein (q.v.). Like judgment’,

GLOSSARY

183

doxa can mean either a cognitive faculty or the results of its operations. The former meaning is the

more prominent here (e.g. 458b10, b12, 459a6).

doxazein, 'to judge’, 'to have in one's judgment’. Used at 458b25 with a direct object, to mean ‘form a judgment about ...', and at 459a8 as a variant for doxa (q.v.), to mean the faculty by which judgments are formed. dunamis, ‘capacity’, ‘ability’, ‘faculty, ‘power’, 'potential(ity). This term is correlative to energeia (q.v.). It is used of a perceptual capacity, in terms of which ‘waking’ is defined, and whose temporary suspension constitutes sleep. Cf. especially 454a8-10. At 455a16 the word is conjoined with ‘common’, to refer to the general sense-faculty or power which attends upon the exercise of all of the special senses, and which has often been equated with ‘common sense' (koine aisthésis). See note 6 on 455a12-b2. dream, see enupnion. eidôlon,

‘image’

(461a15,

462a14,

417,

464a6,

b9,

b14).

Originally

used

of

‘apparitions’ or ‘phantoms’, this term was adopted by the atomists to signify ‘films’ (Latin simulacra), given off by external objects and penetrating the pores of the body, to explain dreams and fantasies in general. Aristotle, while rejecting the theory (464a5-6), himself uses eiddlon of a dream-image (461a15). He also retains the assimilation of having a dream to seeing a reflection, which underlies the theory. Cf., especially, 464b8-16, and see note 54 on 464b5-16. eknoia, ‘unconsciousness’, 'an unconscious state’. This rare word has been translated

thus at 455b6, 456b10, 457b25. In the first of these passages Hett gives ‘derangement’. But there is no reason to associate loss of sensory awareness with mental disorder. Aristotle may be thinking, rather, of trance states, especially at 456b10-12. But at 457b25 the word must refer to a feature of sleep in general. emotion, see pathos. energeia, ‘activity’, 'actualization'; energein, 'to be active', 'to function’. Literally, these words connote 'being on the job', 'being at work', i.e. not lying dormant. In Aristotle they are technical for the exercise of a capacity (dunamis, q.v.) or disposition, as contrasted with the mere possession of it. In the present works, they occur mainly in connection with the functioning of the senses or intellect (e.g. at 454b9, 459227, 460b32, 46125). At 461b13 ‘actually’ is contrasted with 'potentially',

with respect to the presence of movements in the blood. At 461522, the 'actual' senseimpression is the original response to a perceptual stimulus, as opposed to a residue therefrom remaining in the sense-organs. ennoein,

‘to think’, ‘to have in mind'; ennoia, 'thought'. The verb is used at 458b15,

b18, b25, in distinguishing dreams from what ‘we think’, the latter being properly the objects of judgment. "True thoughts', again distinguished from dream appearances, occurs at 462228.

184

GLOSSARY

enupnion, ‘dream’. This is the standard noun in prose, especially in scientific discourse. In Aristotle it has virtually ousted the more literary onar and oneiros, which are common in Plato. Aristotle does not use the adverbial antithesis of onar and hupar, as Plato often does, to mark the difference between dreaming and waking vision. He does use the verb oneiréttein (453b18, 463b12), but he shows a preference

for enupniazein, which occurs three times in the argument of 459a8-22. An important feature of enupnion and enupniazein is that, unlike the English ‘dream’, they have a conceptual link with 'sleep' (hupnos) built into them. See Introduction, 6-8, and (for Aristotle's usage) Kessels, 190-1, 204. ergon, ‘function’, 'job', ‘work’. Used at 454226, a29--30, 032, of the natural function

of a bodily organ or part. This is a key concept in Aristotle's biology, ubiquitous in his accounts of the parts of organisms, and of their contribution to the preservation and well-being of the whole. An organ such as the eye is properly so called in so far as it is capable of performing its function, doing its job. See also energeia. esomenon, future participle of the verb 'to be' (einai), used at 463b29, with the article (to), to mean ‘what will be the case ' as distinct from ‘what is going to be the case’, ‘what is imminent (to mellon, q.v.). eulogos,

reasonable',

‘logical’; eulogôs,

'reasonably'.

The

adjective

is sometimes

given as 'probable'. But the word is used, rather, of what it makes good sense to accept or suppose (e.g. 460a4, 462b19, 46326), or what is ‘intelligible’, as free from contradiction or paradox (cf. 455a31-32, 457b9, 462b8). euthuoneiria, ‘direct dream-vision' (463225, 464b7); euthuoneiroi, ‘persons having direct dream-vision'

(463b16,

464a27).

These

rare words,

found

nowhere

before

Aristotle, may have been coined by him to express the idea of direct or 'straight' apprehension of real events given in dream vision. Cf. EF 1248a40. Anyone can judge such visions (464b7), since their origin is transparent. ‘Clairvoyant’ captures the idea of direct vision, but is confined to paranormal awareness, whereas Aristotle's

term also covers such commonplace dreams as those reflecting the subject's daytime preoccupations (463a23—27). See Introduction, 41. explanation, explanatory factor, see aitia. faculty, see dunamis.

fantasy, see phantasia, phantasma. feeling, see pathos. fortuitous, see automaton. function, see ergon. future, see esomenon, mellon.

GLOSSARY

185

gignesthai, ‘become’, 'come to be’, ‘come about’, ‘happen, ‘occur’. This verb can be used either with a complement (X becomes F') or without one ('X comes to be’). In the latter use, it can signify the coming about of events, and it occurs commonly in this sense, especially in De Divinatione, e.g. at 462b27, 463b5, b11. hupar, ‘waking’, used adverbially in contrast with onar, 'dreaming', though not by Aristotle. Cf. Introduction, p. 8, and see enupnion. hupnos,

'sleep'. When

this noun occurs in the plural (e.g. 458b13,

462b12),

it has

been translated 'periods of sleep'. image, see eidolon, phantasma. imagine, imagination, see phantasia, phantastikon. judge, judgment, see dokein, doxa, doxazein, krinein. kinein, verb, active 'to move' (transitive), to impart movement'; middle, 'to move' (intransitive), 'to move oneself’; passive, ‘to be moved'. The verb, like the noun kinésis (q.v.), has a wider connotation than local movement. It can also mean 'to change’, ‘to modify’. The intransitive use of ‘move’ or 'change' is covered by the Greek

middle

voice, but since this often takes the same

sometimes unclear whether 'moves' sense. At 460b15

form

as the passive, it is

(intransitive) or ‘is moved' gives the intended

the verb has been translated 'move themselves'. At 460b22-27, the

translation '1s itself being moved' may mislead the reader as to the point being made there. See note 35 on 460b16-27 (p. 148). kinésis, noun, movement, 'motion', 'change'. 'Movement' has generally been used for this noun, but is doubly unsatisfactory. (1) The English word is restricted normally to local movement, whereas kinésis can cover other kinds of change. See Plato Theaetetus 181c-d. For a discussion of its definition in general, see Phys. IIL1. In Phys. V. ] Aristotle distinguishes three kinds of kinesis : quantitative, qualitative (alloiösis, q.v.),and local change.

At its broadest, kinesis can cover any change in the

condition of a subject between one time and another. Its wide scope is clearly evident in such contexts as 459a28-b7 or 463b25. Accordingly, kinésis in conjunction with arché (q.v.) has been given as 'source of change’. (2) It is often particularly difficult to render in the plural. Sometimes 'movements' are changes in the sense-organs, occurring during or after perception, and transmitted to or from the heart (e.g. 460b28-29,

461b12,

462a8, 46328). Elsewhere

(46429, a18, a31) they are impulses

propelled from external objects towards a sleeping person. 463b25, Aristotle's usage seems far wider, the word being occurrence that modifies an existing state of affairs. Where it how Aristotle conceived of kineseis, 'movements' renders them degree of vagueness.

Occasionally, as at used to cover any remains unclear just with an appropriate

186

GLOSSARY

krinein, ‘to judge’, ‘to interpret. Sometimes used of judgments based on the suggestion of the senses, or of reason, or both (460b17, b22). At 461b6 a phrase formed on this verb (to epikrinon) signifies the critical faculty, which is commonly inhibited or enfeebled during sleep, so that the dreaming subject is misled. At 464b7 and b12, krinein has been translated ‘interpret’. It there governs ‘direct dream-visions' and ‘reflections’ as its objects, and signifies an activity practised upon the dream by a waking subject, not necessarily the dreamer himself. Just what is meant by ‘interpreting’ dreams in this connection, however, remains unclear. See note 54 on 464b5-16, and Introduction, 41-2. logos, ‘account’, sometimes used of the definition or formula in which a thing's essential nature is expressed. The word occurs in a brief resumé of Aristotle's doctrine of ‘four causes’ at 455b14—16, where it represents the 'formal cause’. luck, see automaton, daimonios. mantiké, used,

with

‘of divination’. This adjective, formed on mantis, ‘the art of (techné)

understood,

‘a diviner' or ‘a seer', is

in the title of the De Divinatione.

The

noun manteia is sometimes given as ‘prophecy’, but 'the mantic art’ is not restricted to foretelling the future, and can cover paranormal vision of present or past events. See note 44 on 462b12-26. melancholic, see next entry. melancholikos,

'atrabilious'. This translation reflects the connection of the term, in

Aristotle's physiology, with black bile (melaina cholé). It refers to a strongly excitable temperament, due to the presence or excess of this substance. For a detailed account of the effects of black bile, cf.[ Probl.] xxx, esp. 954a12-38. The condition is noted as (i) strongly affected by mental presentations (DM 453219), (ii) a cause of poor sleeping and much eating (457a27—33), or of (iii) grotesquely distorted dreams (4612322), (iv) producing unusually numerous and varied dream visions (463b17), and (v) making good guessers (464a32-33). The atrabilious are also said to be impetuous (EN 1150b25-26), and to be constantly in need of pleasure to counteract their strong desires (EN 1154b11--15). See also note 53 on 464a27-b5. mellein, ‘to be going to’, ‘to be about to’. This verb often signifies, as at 463a23-24, a human agent's intention to act as indicated in a further verb. The phrase to mellon or (plural) ta mellonta is used more broadly, to mean ‘the future' or ‘future events’. Cf. 453b22, 463a19, 464a18-19. However, at 463b28-29 Aristotle distinguishes to mellon, 'what is going to be' from to esomenon, 'what will be'. The latter is what will inevitably occur, whereas the former is only an event that will occur unless some factor should intervene to prevent it. At 2:00 p.m. one may be 'going to' take a walk at 3:00 p.m., but by 3:00 p.m. one may have changed one's mind and therefore not take it. If so, then although it was true at 2:00 p.m. that one was 'going to' take the walk at 3:00 p.m., it was not true that one 'would' take it. The gap between what is

GLOSSARY

187

‘going to' happen and what actually ‘will’ happen includes, but is not limited to, the gap between human intention and execution. It is also found, e.g., in the weather. Cf.

463b25-26, and see note 50 on 463b22-31 (pp. 164-5 on b28-29). Aristotle's use of mellon is comparable with the English use of ‘impending’ or ‘imminent’ for prospective events (e.g. a storm, an election) which will not necessarily occur. noétikon,

‘thinking’, ‘intellectual’. This word is used, with the article (fo), to mean

‘the thinking part’. The formation, which is parallel to ‘the perceptual part’, 'the imagining part' (to aisthétikon, to phantastikon, q.v.), is cognate with noein, 'to think, ‘to understand’. It is used at 458b1, synonymously with to doxazon (q.v., 459a8), to signify the capacity to entertain thoughts or make judgments. To ask whether dreams are the work of to noétikon is to ask whether dreaming is, or involves, the having of thoughts or the making of judgments. See also note 20 on 458a33-b3. onar, oneiros, oneiröttein, 'dream'. See enupnion and hupar. opsis, 'sight', vision. This word has been translated 'sight' when it means the faculty of sight (455a17,

419, 458b3,

45923,

460b22),

and ‘vision’, when

it means

visual

experience (459b15, 461a21, 463b18) or gaze (459b13). In one passage, however, (459b27, 460al, 460a3) where it seems to be used synonymously with ‘the senseorgan that perceives colours' (460a25-26), it has been rendered 'organ of sight’. This may be the meaning also at 461a28 and 461b1, though the word has there been

translated ‘vision’. origin, see arché. passion, see next entry. pathos, plural atrabilious, ‘affection’ ‘condition’, ‘state’, emotional state’, ‘feeling’. This noun, cognate with paschein ‘to undergo’, ‘to be affected’, 'to suffer’, has been translated ‘affection’ when it signifies a state, such as sleep or dreaming, in which the subject is thought of as passively affected. In this use there is no suggestion of ‘suffering’, but merely an implication that the subject is undergoing rather than doing something. In a narrower use, however, the word can mean (e.g. at 460b4, b11, b13,

b15) 'emotional state', regarded as a mental perturbation involving pleasure or pain. ‘Passion’ in its older use, e.g. in Hume, reflects the kinship of the emotions with 'passivity', but its modern use is too narrow for the required range of feelings and states. Accounts of the pathé play a major role in Aristotle's ethical theory (see, especially, EN II-IID, and in the Rhetoric (especially Book IT). See also note 34 on

460a32-b16. perceive, perception, see aisthanesthai, aisthésis, aisthétikon. phainesthai, 'to appear’, ‘to be apparent’, 'to be evident’. This verb is part of the wordgroup that includes phantasma phantasia, phantastikon, phantazein (q.v.),

188

GLOSSARY

ancestors of the English ‘fantasy’, 'fancy' etc. Phainesthai, when used with a complement (F), can mean either (1) 'to appear to be F', with no implication that the subject is really F, or (2) 'to be evidently F', with the implication that the subject certainly is F. When phainesthai is followed by an infinitive, the meaning is (1), when by a participle (2). However, where there is no substantive verb, it is not always clear which of these meanings is intended. (1) is appropriate, as Aristotle notes at DA 428a12-15, for occasions on which some item is indistinctly perceived, so that caution is registered in saying merely that it 'appears’ to be F. The speaker recognizes that it may not really be so, and may even know quite well that it is not. Aristotle distinguishes, accordingly, between what 'appears' (phainetai) and what ‘is judged' or 'seems' (dokein, q.v.) to be the case, though he does not always observe this distinction in his own usage. In addition, (3) phainesthai is often used absolutely (e.g. 458b1, 461a15, 427-28, 461b5, 462a7, 462b8), to mean that some item ‘appears’, i.e. there is an appearance of that item to a subject, whether or not there is any such item really present (as in dreams or hallucinations there is not). More generally, the verb signifies ‘phenomena’, i.e. things that 'appear', both in ordinary perception and in non-standard experience, such as dreaming. phantasia, ‘appearance, ‘presentation’, ‘imagination’. This concept, discussed in DA IIT. ch.3, plays a crucial role in Aristotle's account of dreams. The definition of it given at DA 429a1-2 is used at 459a17-18. It is connected variously with (i) 'appearances' in normal waking sense-experience, (ii) memory images, (iii) what appears good or pleasant, and thus as an object of desire, (iv) mental images formed in the course of waking thought, (v) optical and other sensory illusions, (vi) dreams or visions, and (vii) delusions and hallucinations. No one English word signifies a factor common to all of these. In some places phantasia, or to phantastikon (q.v.) used equivalently, seems to stand for a mental faculty or disposition whereby appearances (phantasmata, q.v.) are presented to us. In this use, ‘imagination! is sometimes a possible translation, especially where phantasmata may be interpreted as mental images. But elsewhere (e.g. 456b12, 457b26, 459217), phantasia signifies, rather, the state of the subject when the faculty is being exercised, the condition of being appeared to, or the way an experience registers with the subject at that time. For these uses ‘imagination’ is not at all suitable. Nor need the word contain any reference to mental imagery at all. In still other places (e.g. 460b19-20, 462a8), it is unclear whether a mental

faculty, a mental

state, or a presentation of some

sort is

meant. The word has been variously given as ‘appearance’, or imagination’, or ‘fantasy’, but questions of interpretation arise continually. See Introduction, section 5. phantasma, ‘appearance’. Literally, something ‘made apparent’, ‘made to show’, or ‘presented’. Hence, 'an appearance’, particularly an unreliable or deceptive one. The word occurs in Aristotle's definition of a dream at 462a29 (cf. 458b18, 459a19). In such contexts ‘image’ is a possible interpretation. But its connections with the wider notion of ‘appearance’ are never far from the surface, as is shown especially by

GLOSSARY

189

examples in the context of 460b17. Some of the uses illustrated at 462a15-31 involve no reference to imagery at all. See note 35 on 460b16-27 (pp. 147-8 on b19--20, b20-22), and phainesthai, phantasia, phantastikon, phantazein.

phantastikon, ‘imagining’. Used with the article (fo), to signify a part of the soul whereby 'appearances' (phantasmata q.v.) are presented. The phrase is used equivalently to phantasia (q.v.), where that term signifies a mental capacity or disposition. In some contexts phantasmata may be understood as mental images, and to phantastikon thus corresponds to one aspect of the imagination, viz. the capacity for having or forming images. But elsewhere its scope seems wider. It is the capacity whereby things presented to a subject, in any mode of experience, appear to that subject in the way that they do. At 459a15-17 it is identified with the perceptual faculty (to aisthétikon, q.v.), yet treated as conceptually distinct from it. See Introduction, section 6.

phantazein, verb, active 'to make apparent’; passive, ἴο be presented’, ἴο be conjured up’. Used at 464b1 of an appearance or image presented to mental vision. Cognate with phainesthai, phantasia, phantastikos, phantasma (q.v.). phantom, see eidölon, phainesthai, and Introduction, 6-10. presentation, see phantasia, phantasma, phantastikon, and Introduction, section 6. privation, see sterésis.

psuché, ‘soul’. The factor that distinguishes living from non-living things, sometimes treated by Plato as an agent animating the body in which it dwells. In the present works,

Aristotle

does

not take this dualistic

view.

Rather,

he holds

the position

worked out in the De Anima II. 1-3, according to which soul and body are not separate substances, but complementary aspects of a single living thing. On this view, a soul can no more exist independently of a body than a capacity for 'chopping' (to use an Aristotelian example from DA 412b10-413a2) can exist independently of an axe whose capacity it 1s. For any living thing, the soul is, roughly, the set of capacities in whose exercise that thing realizes its essential nature. Accordingly, sleep is ascribed

neither to the soul nor to the body

whole (453b25-454a11). is ill suited

to Aristotle's

scientific approach. alternatives.

alone, but to the animal

as a

The word 'soul', with its dualistic and religious overtones, subtle

view

of the mind-body

But it is traditional and no more

reason, see aitia. result, see sumbainein. seem, see dokein, phainesthai.

sense, sense-faculty, see aisthanesthai, aisthesis.

relationship,

misleading

and

to his

than possible

190

GLOSSARY

sense-impression,

sense-datum, see aisthésis, aisthéma.

sense-object,

sensum, see aisthéma, aisthéton.

sense-organ,

sensorium, see aisthétérion, aisthétikon.

sight, see opsis. sleep, see hupnos. soul, see psuché.

source, see arche. starting-point, see arché, and note 30 on 459a28-b7 (pp. 144—5 on 459b3).

sterésis, 'privation'. The lack or absence of an attribute, particularly, though not necessarily, one that the subject would normally possess by nature. Blindness, for example, is a privation of sight. For a general account of the concept, see Metaph. V. 22. Sleep is called 'a sort of privation of waking at DS 453b26. The qualification suggests that it does not fit exactly into any of the categories of sterésis distinguished in the Metaphysics. Unlike blindness or deafness, it is not contrary to an animal's normal nature. Nor is it a lack of any power that an animal ought to have, but rather a temporary suspension of those it does have. sumbainein, ‘to happen', 'to result’, 'to follow’. This verb sometimes merely means ‘to

happen’. But it can also cover ‘consequences’, of either a factual or a logical kind, i.e. happenings consequent upon other events or conclusions consequent upon other propositions. It has been variously translated, according to context, ‘in effect’, ‘as a result’, ‘in consequence’, ‘there have been cases of ...’.

sumptóma, 'coincidence'. Aristotle's term for two events so related that neither is a 'sign' or a 'cause' of the other. By an odd reversal, its modern derivative, 'symptom', means, in medical contexts, what Aristotle calls a 'sign'. Cf. 462b31-463a3,

note 45 on 462b26-463a3. think, thought, see doxa, doxazein, ennoein, krinein, noétikon. unconsciousness, see eknoia. vision, see opsis.

and see

191

Select Bibliography Although the essays of the Parva Naturalia can be read independently of Aristotle's other writings, they are best studied against a wider background. The select bibliography is limited to works, mainly in English, that will be reasonably accessible to modern readers. It has been updated to include a number of studies omitted from, or published since, the first edition of this book. The short book by Ackrill contains a lucid overview of Aristotle's thought, and is an excellent guide to his work as a philosopher. Chapters 4—5, on explanation in natural science and the philosophy of mind, are particularly relevant to the subjectmatter of this volume. The concise introduction by Barnes is also highly recommended. An outline of Aristotle's life, with a useful overview of his output, can be found in

Barnes 1995, which also includes a wide-ranging bibliography. The great work of Dodds, especially chapter 4, is invaluable with regard to linguistic matters and literary background. Other fine studies, of the first importance for topics relevant to the Parva Naturalia, include those of Block, Cooper, Kahn, Lowe, Nussbaum, Owens, Schofield, and Sorabji.

The succinct Latin commentary by

Siwek is often helpful, as are the notes to Beare's translation (1908). I have not yet seen the edition of De Insomniis and De Divinatione per Somnum by van der Eijk. The fullest philosophical discussion of dreams to have survived from antiquity is that of Cicero, which is well worth reading.

The Oneirocritica

by Artemidorus

of

Daldis, dating from the second century A.D., is a rich and entertaining source for the ancient art of dream interpretation. There is a vast and growing contemporary literature on dreams. The bibliography contains only a small mentioned in this book.

fraction of it, but includes all modern sources used or In addition, some influential remarks about dreams will be

found in Wittgenstein. Dunlop has edited a useful anthology of recent philosophical writing on dreams, much of it provoked by the work of Malcolm. Many penetrating observations are also to be found in the philosophical essays of Hunter. Notable recent works in contemporary experimental psychology of dreaming are those of Hobson, Smith and Winson.

Ackrill, J.L. Artemidorus

Aristotle the Philosopher. Oxford: 1981. The Interpretation of Dreams.

Oneirocritica, trans. R. White, Park

Ridge, N.J.: 1975. Athenaeus

Deipnosophistae. trans. C.B. Gulick. Loeb Classical Library (7 vols), vols. IT, IV. Cambridge, Mass: 1957.

192 Barnes, J.

BIBLIOGRAPHY ed. with M. Schofield and R. Sorabji, Articles on Aristotle, vol.1

(Science).

London: 1975.

ed. with M. Schofield and R. Sorabji, Articles on Aristotle, vol. 4

(Psychology and Aesthetics). Aristotle.

Oxford:

London: 1979.

1982.

ed. The Complete Works of Aristotle. Revised Oxford Translation, 2 vols. Princeton: 1984. ed. The Cambridge Companion to Aristotle, Cambridge: 1995. Beare, 1.1.

Greek Theories of Elementary Cognition. Oxford: 1906. trans. De Sensu et Sensibilibus, De Memoria et Reminiscentia, De Somno et Vigilia, De Insomniis, De Divinatione per Somnum in Oxford

Translation

of the Works

of Aristotle,

vol.III.

Oxford:

1908. Benedetti, F.

"Processing of Tactile Spatial Information with Crossed Fingers.' Journal of Experimental Psychology: Human Perception and Performance, 11 (1985): 517-235.

Biehl, W.

ed. Parva Naturalia.

Block, I.

"The Order of Aristotle's Psychological Journal of Philology 82 (1961): 50-77.

Bos, A.P.

'Manteia in Aristotle, De Caelo 11.1', Apeiron

Burnyeat, M.F.

‘Idealism and Greek Philosophy: What Descartes Saw Berkeley Missed.‘ Philosophical Review, 91 (1982): 3-40.

Cicero

De Divinatione, trans. W.A. Cambridge, Mass:

Cooper, J.M.

Leipzig: 1898. Writings.’ American 21 (1988), 29—54.

|

and

Falconer. Loeb Classical Library,

1923.

‘Aristotle on Natural Teleology.' Language and Logos, eds. M.C. Nussbaum and M. Schofield. Cambridge:

Crick, F. and Mitchison, G. 14.

1982, 197—222.

"The Function of Dream Sleep.’ Nature 304 (1983): 111—

Croom Robertson, G. Note in Mind 1 (1876): 145-6. Devereux, G.

Dreams in Greek Tragedy. Oxford: 1976.

Diels-Kranz.

Die Fragmente der Vorsokratiker. DK).

Diogenes Laertius

The Lives of the Philosophers, trans. R.D. Hicks. Library (2 vols). Cambridge, Mass: 1959.

Dodds, E.R.

The Greeks and the Irrational. Berkeley: 1951.

Drossaart Lulofs, H.J.

Berlin: 1951 (abbreviated as Loeb Classical

ed. De Somno et Vigilia. Templum Salamonis: 1943.

BIBLIOGRAPHY

193

ed. with preface De Insomniis et De Divinatione per Somnum, 2 vols. Leyden: 1947. Dunlop, C.E.M.

ed. Philosophical Essays on Dreaming. Ithaca: 1977.

Edel, A.

Aristotle and his Philosophy. Chapel Hill: 1982.

Evans, C.

Landscapes of the Night. ed. and completed, P. Evans. London:

1983. Freud, S.

The Interpretation of Dreams, vols.IV-V of the Complete Psychological Works of Sigmund Freud. trans. and ed. J. Strachey with A. Freud, A. Strachey, and A. Tyson. London:

Gallop, D.

1955.

‘Dreaming and Waking in Plato.’ Essays in Ancient Greek Philosophy, eds. J. Anton and G.L. Kustas. Albany: 1971, 187-201. trans. and comm. Plato, Phaedo. Oxford: 1975. ‘Aristotle on Sleep, Dreams, and Final Causes.’ Proceedings of the Boston Area Colloquium in Ancient Philosophy, vol IV (1988), eds. J.J. Cleary and D.C. Shartin, Lanham: 1989, 257-90.

Greek Anthology.

trans.

W.R.

Paton.

Cambridge, Mass:

Loeb

Classical

Library

(5

vols),

vol.

IE.

1960.

Grube, G.M.A.

trans. Plato, Republic. Indianapolis: 1974.

Hamlyn, D.W.

trans. and comm.

Aristotle,

De

Anima,

Books

II-III. Oxford:

1968. Hankinson, R.J.

‘Aristotle on Sleep and Dreams: A Text and Translation.’ Dialogue 33 (1994): 340-3 (review of first edition of the present work).

Hardie, W.F.R.

‘Aristotle's Treatment

of the Relation

between

the Soul and the

Body. 'Philosophical Quarterly 14 (1964): 53-72. Hett, W.S.

trans. Aristotle,

Classical Library.

On the Soul, Parva Naturalia,

On Breath. Loeb

Cambridge, Mass: 1936.

Hippocrates

On Regimen. trans. W.H.S. Jones. Hippocrates (4 vols), vol IV. Loeb Classical Library. Cambridge, Mass: 1931.

Hobson, J.A.

The Dreaming Brain. Basic Books, New York: 1988.

Huby, P.M.

"The Paranormal in the Works of Aristotle and his Circle; Apeiron

13 (1979): 53-62. Hunter, J.F.M.

"Mullane on Responsibility for Dreams.' Dialogue 4 (1965-66):

531-5. ‘Some Questions about Dreaming. Mind 80 (1971): 70-92. Reprinted in Essays After Wittgenstein, Toronto: 1973, 67—89.

194

BIBLIOGRAPHY "A Puzzle about Dreaming.’ Analysis

(1976): 126-31.

"Ihe Difference between Dreaming and Being Awake.’ Mind 92

(1983): 80-93. Jackson, H.

‘Aristotle's Lecture-Room and Lectures’, Journal of Philology 35 (1920): 191—200.

Kahn, C.H.

‘Sensation and Consciousness in Aristotle's Psychology.’ Archiv für Geschichte der Philosophie 48 (1966): 43-81. References are to the reprinted version in Barnes 1975, 1-31.

Kessels, A.H.M.

Studies on the Dream in Greek Literature. Utrecht:

Lesher, J.H.

"Xenophanes' Scepticism.’ Phronesis 23 (1978):1-21.

Lloyd, AC.

1978.

"Ihe Principle that the Cause is greater than its Effect.’ Phronesis

21 (1976): 146-51. Lloyd, G.E.R.

"The Empirical Basis of the Physiology of the Parva Naturalia’. Aristotle on Mind and the Senses. eds. G.E.R. Lloyd & G.E.L. Owen. Cambridge: 1978, 215-39.

Lowe, M.

'Aristotle's De Somno and his Theory of Causes.'

Phrenesis 23

(1978): 279-91. Lucas, D.W.

ed. and comm. Aristotle, Poetics. Cambridge: 1968.

Lucian

trans. A.M. Harmon, M.D. Macleod. Loeb Classical Library (8 vols), vols V, VIII.

Cambridge, Mass:

1962, 1967.

Lucretius

De Rerum Natura. trans. and comm. C. Bailey.

Malcolm, N.

Dreaming. London: 1959.

Modrak, D.

Aristotle: The Power of Perception. Chicago: 1987.

Mourelatos,

A.P.D. ‘Aristotle's Rationalist Account Phronesis 29 (1984): 1-16.

Oxford: 1947.

of Qualitative

Interaction.'

Mugnier, R.

ed. with introduction Parva Naturalia. Paris: 1953.

Nussbaum, M.C.

ed. and comm. Aristotle, De Motu Animalium. Princeton: 1978.

Owens, J.

"The Teleology of Nature in Aristotle.’ The Monist 52 (1968), 159—73. repr. in The Collected Papers of Joseph Owens, ed. J.R.Caton, Albany: 1981, 136-47, 216-18.

Peck, A.L.

trans. Aristotle, Historia Animalium, (3 Classical Library. Cambridge, Mass. 1970.

Pliny

Natural History, ttans. H. Rackham. Loeb Classical Library (10 vols), vol II. Cambridge, Mass:

Plutarch

vols),

vol

II.

Loeb

1947.

Quaestiones Conviviales, trans. H.B. Hoffleit, E.L. Minar. Loeb Library (15 vols), vols. VIII, IX. Cambridge, Mass:

Classical

1961,1969.

BIBLIOGRAPHY

195

The Histories, trans. W.R. Paton. Loeb Classical Library (6 vols),

Polybius

vol. IV. Cambridge, Mass:

1960.

‘On Dreams II. 459b24—460233, and Aristotle's opsis.' Phronesis

Preus, A.

13 (1968): 175-82. Putnam, H.

"Dreaming and "Depth Grammar", Butler, Oxford: 1962, 211—35.

Analytical Philosophy, ed. RJ.

Ross, W.D.

ed. and comm. Parva Naturalia. Oxford: 1955. ed. and comm. De Anima. Oxford: 1961. 'A New

Saver, J.

Interpretation

of Dreams',

London

Review

of Books,

August 4, 1988: 11-12 (review of Hobson).

Schofield, M.

'Aristotle on the Imagination.' in Aristotle on Mind and the Senses, eds. G.E.R. Lloyd and G.E.L. Owen. Cambridge: 1978, 99—129. References are to the reprinted version in Barnes 1979,

Sextus Empiricus

Opera, trans. R.G. Bury. Cambridge, Mass: 1968.

Siwek, P.

ed. Aristotelis Rome: 1963.

103-32.

Smith, C.

Parva

Loeb

Classical

Naturalia

Library

(4

vols).

with Latin trans. and comm.

‘Sleep States and Learning: A Review of the Animal Literature." Neuroscience and Biobehavioral Reviews, 9, (1985);

157—68.

Aristotle on Memory. London: 1972.

Sorabji, R.

Necessity, Cause and Blame: Perspectives on Aristotle's Theory. London: 1980. Sprague, R.K.

‘Aristotle and the Metaphysics of Sleep.' Review of Metaphysics

31 (1977): 230-41. ‘Aristotle on Red Mirrors Phronesis 30 (1985): 323-5.

Dreams

‘Plato's Philosophical Uses of Dream Journal of Philology 91 (1970): 204-12.

Tigner, S.S. van der Eijk, P

(On

J.

II 459b24-460a23).' Metaphors.’

American

trans. and comm. Aristoteles, Uber Traiime; Uber die Weissagung im Schlaf. Berlin: 1994.

van Groningen, B.A.

Mnemosyne, ser. 4. (1) (1948), 107-8.

van Lieshout, R.

Greeks on Dreams. Utrecht:

Walbank, F.W.

A Historical Commentary 1967.

1980.

on Polybius (3vols), vol II. Oxford:

196

BIBLIOGRAPHY

Wieland, W.

"The Problem version

of Teleology.' in Barnes

of ch.

16, ‘Zum

Physik, Gottingen: 1962, Wiesner, J.

"Ihe Unity

1975,

Teleologieproblem',

141-69. English Die

aristotelische

2nd ed. 1970.

of the Treatise De

Somno

and

the Physiological

Explanation of Sleep in Aristotle.' Aristotle on Mind and the Senses.

eds.

G.E.R.

Lloyd

&

G.E.L.

Owen.

Cambridge:1978,

241-80. Wijsenbeek-Wijler, H.

Aristotle's Concept of Soul, Sleep and Dreams. Amsterdam:

1978. Williams, C.J.F.

trans. and comm. Oxford: 1982.

Aristotle,

Winson, J.

"Ihe Meaning of Dreams',

De

Generatione et

Corruptione.

The Scientific American, NS 263 (Nov.

1990): 86-96. Wittgenstein, L. -

Philosophical Investigations. Oxford 1953. Lectures and Conversations on Aesthetics, Religious Belief. ed. C. Barrett. Berkeley: 1967.

Woods, M.J.

‘Aristotle on Sleep and Dreams’, Apeiron (review of first edition of the present work).

Yates, F.

The Art of Memory. London: 1966.

Psychology 2 (1992):

and

179-88

197

Index The figures in bold type refer to the text and translation. Academy 1 activity 33-4, 121--3, 126, 128, 134, 151, 180, 183 Aeschrion 169 after-images 459b7-18, 38, 145, 153 Agamemnon 7 Alexander the Great 1-2 alteration 459b1-5, 181 (see also change, movement) Amyntas IT 1 analogy 25, 42 Aphrodite 464b3, 169—70 apparition

8-10,

152,

183

appear, appearance 456b12-17, 458b15-25, 459a7-8, a19-20, 460b19-20, 461a18, 462a15b8, 463a21-30, 464218, 464b9, 140-54, 181, 187-9 (see also imagine) Aristophanes 7n Aristotle passim Categories 64n De

Anima

4-5,

23-8,

31,

34,

121-2, 124-6, 136, 139, 141, 143, 147, 180, 188-9, 193-5 De Divinatione per Somnum passim De Generatione

Animalium

5, 52,

122, 130, 132, 145-6, 154 De Generatione et Corruptione 128, 133, 164, 196 De Insomniis passim De Interpretatione

64n. 162, 165

De Juventute et Senectute 4 De Longitudine et Brevitate Vitae 4

De

136,

139,

151, 180, 186, 192, 195 De Motu Animalium 116n De Partibus Animalium 5, 128-9, 132-3, 180-1 De Respiratione 4

Memoria

124,

De Sensu

192

38,

4,

124,

122, 125,

De Somno et Vigilia De Vita et Morte 4 Eudemian Ethics

136,

passim

14, 32

Historia Animalium

| v,

35,

51,

123, 155, 173-9 Metaphysics 17-19, 33, 133, 135, 148, 181, 190 Nicomachean Ethics 12n, 32, 43, 122-4, 128, 132, 159, 167, 169 Parva Naturalia 1x, x-xi, 4—6, 17, 116n, 192-5 . Physics 31, 44—5, 120, 128, 144, 158, 163, 181, 185 Poetics

40, 45, 45n,

Posterior Analytics [Problemata]

35,

194

181 51,

130,

132,

148, 186 Rhetoric

44n, 98n,

187

Topics 139 Artabanus 8—9 Artemidorus 191 Assos 1 Athenaeus 169, 191 atomism 464a1—19, 166 atrabilious 457a27, | 461a22, 463b17, 464a32-33, 186 being 454a19, 45521, 23, 121, 125, 143

459a16,

198

INDEX

belief 15, 24, 27 (see also judgment) Berkeley 49 blood 458a10-25, 459b23460a23, 461b11-30, 100n, 1334, 146, 150-2, 183 Borysthenes 462b25, 157 brain 457b26-458a10, 29, 125, 133-4 breath 455b34-456a24, 129 capacity 454a8-19, 455a12-b2, 121, 124-6, 151, 182-3 cause

455b13-28,

463a3-31,

20, 30, 33, 42n, 126-9, 66, 170—1, 181 Chalcis 1-2 chance

44,

158,

5,

133-4,

157-

(see

also

161,

coincidence, fortuitous, luck) change 455b15, 459a28-b7, 463b25-28, 144, 181, 185 (see also alteration, movement) Cicero 3,9, 47n, 163, 191-2 clairvoyance 160, 184

Descartes digestion 2

18 456a30-458a32,

Diogenes Laertius Diotima 47

ii, 1-2, 15, 34, 192

divination 462b12-464b18, 14, 44n, 47, 156-8, 161, 170, 186 Dnieper 157 Don Giovanni 45 double vision 461b30-462a3, 152 dreams passim day-dream 11 direct dream vision 463b16, 463a25, 464b7, 464a27, 41-2, 184 dream interpretation 14, 40-2, 170-1, 186, 191 dream objects

6-8, 55-7

dreaming as symbol for belief 15 dreams of animals 463b12, 536b25-31 (=174-5), 5, 9, 35-6, 43, 51-2, 162-3 dreams of children 461a10, 462a14-15, 462b4, 537b14-15 (=178-9), 35 dreams within dreams 139-40 meaning in dreams 30

coherence 98n, 149 coincidence 462b26-463a3, 463a31b20, 464a4-5, 44, 47-8, 157-9, 161-2, 164, 190 colour 455a22, 458b6, 459b723, 460a25-26, 125, 136, 181, 187

sexual dreams telling dreams

coma

veridical dreams

123

common

sense,

sensibles

daemon, daemonic

453b23,

9, 14, 43-8, 119-20, decaying sense delusion

Democritus

463b14,

163, 182

20, 23, 37-8, 144-6

22, 27, 140-1,

464a5,

166

188

private world of dreams

8, 8n, 18

11, 40 49 8n,

11,

19,

119,

156, 158, 161-4

455a12-

b2, 458a33-b9, viii, 19-20, 124— 6, 136-7, 147, 180, 183 Coriscus 461b23-26, 462a5-6, 10, 18, 24, 27, 42, 55-6, 151-2, 170

37, 130-

dualism

13n, 121,

189

emotion 460b3-16, 24, 26-7, 42, 53, 183, 187 epilepsy 457a7-11, epiphenomena 37-8 Er 131 esse est percipi 49 Euboea 1

461b7-8, 12, 147, 150, 159, 132

INDEX examples in philosophy 25n exhalation 456b17-458a10, 131-2 explanation 455b14-16, 14, 28-38, 126-8,

181, 184 (see also cause)

fable 25n fainting 455b5-6, 456b10-17, 20, 74n, 126, 131 false 18-19, 154 fantasy 456b12, 457b26, ix, 11, 183, 188 (see also imagery, imagination) fatalism

162, 165

fatigue 456b34-35, 462b29-30, 132, 159 fear 460b3-16, 462a12-15, 24, 26-7, 147, 152-3 fiction 10-12, 55-7 flavour 463a14, 458b6, 462220, 463a13-15, 137 (see also taste) food 456a30-458a25, 461a8-14, 19, 35, 37, 130, 132-3 foreknowledge 119, 161 (see also precognition) form 19 Forms 15 fortuitous

453b24,

44 6,

181-2, 184 Freud 11, 28-9, 38-48, 163, 171, 193 frogs 461b15-16, 150 function

454a24-32,

184, 186-7 future 453b22, 164—5, 184

119-20,

139,

154,

199 43, 46-7,

156, 163 (see also daemon,

divination) Greek Anthology 169 guessing 464233, 14, 41, 44n, 186 hallucinations,

122,

463b22-31,

119,

10,

118-19

10n,

(see also

illness, medicine)

hearing 455a15, 458b3, 459a10, 459b21-22, 461bl, 462a2425, 125, 135, 153—5 (see also sound) heart 455b34—456a24, 458a1521, 19-20, 25, 118, 125, 129, 1334, 145, 149—50, 181 Heracles, Pillars of 462b24, Heraclitus 8, 18, 120

157,

167

Hermeias 1 Herodotus 8 Herpyllis 2 Hippocratic treatises

7n,

9,

13n,

30n,

159, 193 Hippolytus 47 Hobbes 20 Homer 7, 7n, 10-11, Howe, Elias 48

13

Hume 187 hylemorphism viii, 19-20 hypnagogic, hypnopompic

462a8-15, 32-8,

hallucinoid

12, 16, 53, 188 health 30-2, 38-9,

169,

illness

457a3-5,

460b11-16, 463a3-21,

visions

152 | 458b25-29,

461a21-25, 30-2,

140,

147,

159,

164 (see also health, medicine)

gamblers 463b20—22, 161, 163 Gates of Ivory and Horn 11 Gibraltar 157 God, gods 462b20-24, 463b1222, 464a19-22, 7, 11, 13-14, 29,

illusion 458b29-459a8, 459b720, 460b3-27, 461b7-462a8, 12, 16, 22-5, 38, 53, 140-1, 147-8, 150-2 imagery, images 461a14-25, 462a15-19, 463b31-464a19,

200

INDEX

464b5-16, 38-9,

53,

6, 9-11, 183,

188

13-16, (see

also

23-6, after-

images, imagination) imagination 458b25-459a22, 460b3-27, 461b7-11, viii, 4, 21-8, 33, 57, 141-4, 147-8, 150-1, 185, 188-9 impending events 165, 186—7 incubation 31-2 insane

464a25—b5,

170,

183

intellect. 461a1, 33-4, 43, 128, 163, 187 (see also judgment, thought) judgment 458b10-459a9, 460b327, 461a25-462a8, 15, 21-8, 523, 94n, 135-42, 147-8, 150-2, 182— 3, 186 (see also thought) know, knowledge

late dreamers 155 Lesbos 1 Libya 17 like, likeness

15, 18, 135

462a31-b11,

34-6,

metaphor

455b21,

25n, 40, 70n

mirrors 459b23-460a23, 14, 145-6 misperception 458b32, 18, 18n Mitylene 1 Mitys 45-6 mnemonic systems 458b20-25, 12n, 26, 139-40 models 25n Moliére

24

movement, motion

453b31-454a11,

455b13-456a29, 459428-460a23, 460b22-462a31, 463a8-31, 463b18-464b16, 13, 20, 23, 39-40, 40n, 129-30, 144-5, 147-53, 154—5, 164—9, 185 (see also alteration, change)

Mysia 1 myth 25n 461b7—30,

10,

151-2 (see also resemblance) Lucian 169, 194 luck 44, 161, 163-4, 168-9, (see also daemonic, fortuitous) Lucretius 9, 52n, 166, Lyceum 2-3, 5, 173

14—16,

Napoleon

55

narrative 6-10, 12, 50, 52, 56 naturalism 8, 14, 163

181-2

194

Macedon 1-2 madness, see insane Malcolm 48-57, 119, 149, 191, 194 matter

memory 453b18-20, 456a24-29, 458b18-25, 36, 49-51, 119, 13940 menstruation 459b23-460a23, 145— 6

19, 132

mechanical explanation 28 Medea 47 medicine 463a3-21, 30, 159 melancholics 42 (see also atrabilious)

necessity

157, 162

Nicomachus 1-2 non-dreamers see late dreamers nutrient part of the soul 454a11-19, 121 Odysseus 55 opposites 454a31-b9, Orestes 47

pain

454b30,

paralysis

120

461a2

27, 132

paranormal cognition 157, 186 pathological states 460b3—16,

147

INDEX perception 455a4-b13, 455b34456a29, 456b17-19, 458a33b15, 459a8-b23, 460a23462a8, viii, 5, 13, 16, 19-25, 28, 33-5, 37-40, 40n, 53-5, 118-26, 129-30, 135-8, 140-5, 147--55, 157, 180-1,

187

(see also

primary

phantom

8-9,

183,

8n,

189

11-13,

25n,

131,

18,

18n,

193 Sophist

13-14,

Statesman Theaetetus

16, 46

25n — 11n,

16,

138, 185 Timaeus 13-14, 46, 157 pleasure 454b30, 455b19, 461a2, 128 Pliny, the elder

(see

also

Queen, the 6, 158 Queen Victoria 42

2

Philaenis 464b2, 169-70 Philip of Macedon 1 philosophy of mind 19-20 picture 12n, 14, 16, 56 ‘place’ system 139-40 Plato 1-2, 7, 8n, 11-19, 25n, 28, 44, 120, 184, 189, 193 Charmides 11 Meno 170 Phaedo 7, 120, 131, 193 Republic

prophecy 47-8, 186 prediction) Protagoras 16-17 psychoanalysis 38-42 purpose 29-32, 46-8 Pythias 1

sense-

organ, sense)

Peripatos, Peripatetic school Phaedra 47 Phaestis 1

201

145, 194

Plutarch 166, 194 Polybius 169, 195 precognition 43-4, 119, 156, 158, 161, 167-70 prediction 30, 42n, 47-8, 160-2 (see also foreknowledge, precognition) primary sense-organ, -faculty 455a20b2, 455b10-13, 19-20, 124-6, 129, 142, 145 privation 453b26, 120, 189-90

rationalism 8-10 reflections 461a14-25, 464b5-16, 14-16, 55-6, 149, 170-1 REM sleep 10n, 35-7, 57 resemblance

460b3-16,

170 rest 455b18-22,

40-1,

147,

70n

secondary elaboration 10, 10n, 139-40 seeing 454a28,455a7-27, 458b39,458b14-15,458b33-459a14, 4597-20, 459b23-460a23, 460b3-16, 461a25-462a15, 462a21-24, 463b15-18, 7, 7n, 8, 22, 38, 52-3, 124—5, 135-8, 1412, 145-8, 152-3, 187, 190 seeing as

460b3-27,

461b19-21,

23-5, 147-8, 151 seem 182, 188 (see also appear, judgment) sense(s) 19-22, 25, 33-4, 37-8, 52, 124-6, 129, 135-8, 141-2, 148, 152-3, 180-1 (see also perception) sense-faculty 5, 124-5, 180, 189 (see also common sensefaculty) sense-impression

129, 146-7, 190 sense-object

21,

151-5,

136-7,

52,

123,

168,

180,

181, 190

202

INDEX sense-organ

20,

148-53,

129-30,

180,

190

145-6,

(see

also

(see

also

primary sense-organ) sense-remnant

150-3

decaying sense) special senses viii,

5,

124-6,

135-7, 141-2 Sextus Empiricus 13n, 166, 195 sight 152, 187, 190 (see also seeing)

signs 462b26-463a31, 463b2231, 30, 119, 157-9, 164-6 simile 25n sleep, passim (see also REM sleep) actions in

456a24,

130

taste

455a17-21,

108n,

160

(see

also flavour) teleology 14, 28-38 telepathy 464a27-32, 167-8 Theaetetus 16 thought 458a33-b29, 462b27, 910, 51, 53, 135, 137-8, 153-5, 183, 187 (see also intellect, judgment) throttling 455b6-8, 456b10, 126 touch

455a20-27,

460b16-22,

23-4, 126, 148, 152 true 16-17 truth 455b20-21, 70n truth-value 153-4

deprivation of 34 learning in

uncanny

29, 36

sleep-walking, see somnambulism sounds heard in 463a12-13, 38-9, 153-4 smell 459a21-22, 460a26-32, 126, 154 Socrates 2, 11, 47, 55-6 somnambulism 17, 129-30, 154 soporifics 456b28-457a3, 457b6-

26, 132 soul

453b11-454a19,

— 462a6,

supra-human

agency

120

12, 38-42,

170

daemon) symbolism

daemon,

vision

28, 40-2, 46-8

7-8 (see also seeing)

waking, passim waking as perceiving

454411,

waking as symbol or activity

453b24-

120-2 for knowledge

15, 34, 34n

waking as 'the end'

455b22-28,

waking up 458a10-25, 134 wave theory 463b31-464b5, 167-8

will 26-7, 31-2, 160-1, 166-7 wish-fulfilment 11-12, 12n, 28, 30n, 42, 47, 47n

6-7, 12, 27-8, 48, 56

163

(see also

33-4, 128, 134

153-4 (see also hearing, sleep) spirit 47 (see also daemon) Stagira 1 subjectivism. 17 subjectivity 8, 8n supernatural 43, 120, daemon)

163

Unconscious, the

464a10, 4-6, 12-14, 19, 120-1, 189—90 sound 455a22, 458b6, 459b2021, 462a24-25, 38-9, 125, 136,

story

46-8,

fortuitous) unconscious states 455b2-13, 456b9-28, 457b20-26, 19-21, 49, 54, 122, 126, 183

(see also

Wittgenstein

191, 196

Xenophanes Xerxes 8

8, 170, 194

(see

also

Zeus

7

ARIS ἃ PHILLIPS CLASSICAL TEXTS Published and new books

AESCHYLUS EUMENIDES ed. A.J. Podlecki (1989) PERSIANS ed. E. Hall (1996) ARISTOPHANES ed. Alan H. Sommerstein ACHARNIANS (1980) BIRDS (1987) CLOUDS (1982) FROGS (1997) KNIGHTS (1961) LYSISTRATA (1990) PEACE (1985) THESMOPHORIAZUSAE (1994) WASPS (1983)

ARISTOTLE ON THE HEAVENS I & II ed. S. Leggatt (1995) ON SLEEP AND DREAMS ed. D. Gallop (1996) AUGUSTINE SOLILOQUIES (1990)

and IMMORTALITY

OF

THE

SOUL

ed. G. Watson

CAESAR CIVIL WAR I & II ed. J.M. Carter (7991) CIVIL WAR III ed. J.M. Carter (1993) CASSIUS DIO ROMAN HISTORY Books 53.1-55.9 ed. J.W. Rich (7990) CATULLUS 61—68 ed. J. Godwin (7996) CICERO TUSCULAN DISPUTATIONS I, ed. A.E. Douglas (1985) TUSCULAN DISPUTATIONS ἢ & V ed. A.E. Douglas (1989) ON FATE with BOETHIUS CONSOLATION V ed. R.W. Sharples (7992) PHILIPPICS II, ed. W.K. Lacey (1986) VERRINES I1,1 ed. T.N. Mitchell (7956) ON STOIC GOOD AND EVIL ed. M.R. Wright (1991)

LAELIUS ON FRIENDSHIP and THE DREAM Powell (1991) LETTERS (43 B.C.) ed. M. Willcock (7996)

OF SCIPIO

ed. J.G.F.

EURIPIDES ALCESTIS ed. D. Conacher (1988) ANDROMACHE ed. M. Lloyd (1995) BACCHAE ed. R. Seaford (1996) ELECTRA ed. M.J. Cropp (1988) HECUBA ed. C. Collard (1991) HERAKLES ed. S. Barlow (1996) HIPPOLYTUS ed. M.R. Halleran (1995) ION ed. K.H. Lee (1997) ORESTES ed. M.L. West (7987) PHOENICIAN WOMEN ed. E. Craik (/986) TROJAN WOMEN ed. S. Barlow (1986) FRAGMENTARY PLAYS VOLUME I eds C. Collard, M.J. Cropp & K.H. Lee

GREEK ORATORS I ANTIPHON, LYSIAS ed. M. Edwards & S. Usher (7985) II ISOCRATES PANEGYRICUS and TO NICOCLES ed. S. Usher (7990) IV ANDOCIDES ed. M. Edwards (1995) V VI

DEMOSTHENES APOLLODORUS

HELLENICA

On the Crown ed. S. Usher (1993) Against Neaira ed. C. Carey (1992)

OXYRHYNCHIA

ed. P.R. McKechnie & S.J. Kern (1988)

HOMER ODYSSEY I & II ed. P.V. Jones (7991) ILIAD VIII & IX ed. C.H. Wilson (1996) HORACE SATIRES

I, ed. P. Michael Brown (7993)

SATIRES Il ed. Frances Muecke (1993) JOSEPH OF EXETER THE TROJAN WAR I-II ed. A.K. Bate (7986) LIVY Book Book Book Book Book

XXXVI ed. P.G. Walsh (7991) XXXVII ed. P.G. Walsh (1992) XXXVIII ed. P.G. Walsh (1994) XXXIX ed. P.G. Walsh (1995) XL ed. P.G. Walsh (1996)

LUCAN CIVIL WAR VIII ed. R. Mayer (1981) LUCIAN A SELECTION ed. M.D. McLeod (1991) LUCRETIUS DE RERUM NATURA IV ed. J. Godwin (1987) DE RERUM NATURA VI ed. J. Godwin, (1991) MARTIAL EPIGRAMS V

ed. P. Howell (1996)

MENANDER SAMIA ed. D.M. Bain (1983) THE BAD-TEMPERED MAN ed. S. Ireland (7995) OVID AMORES II ed. 1. Booth (1991) METAMORPHOSES LIV ed. D.E. Hill (7985) METAMORPHOSES V-VII ed. D.E. Hill (1992) PERSIUS THE SATIRES

ed. J.R. Jenkinson (1981)

PINDAR SELECTED ODES ed. S. Instone (7996) PLATO MENO ed. R.W. Sharples (7985) PHAEDRUS

ed. C.J. Rowe (1986)

REPUBLIC V ed. S. Halliwell (7993) REPUBLIC X, ed. S. Halliwell (1988) THE STATESMAN ed. C.J. Rowe (1995)

PLAUTUS BACCHIDES ed. J.A. Barsby (1986) PLINY CORRESPONDENCE Williams (1990)

WITH

TRAJAN

FROM

BITHYNIA

PLUTARCH LIVES OF ARISTEIDES AND CATO, ed. D. Sansone (1989) LIFE OF CICERO ed. J.L. Moles (1989) MALICE OF HERODOTUS ed. A.J. Bowen (1992)

ed.

W.

THE RUODLIEB

ed. C.W. Grocock (1985)

SENECA LETTERS: A SELECTION ed. C.D.N. Costa (1988) FOUR DIALOGUES, ed. C.D.N. Costa (1994), SOPHOCLES ANTIGONE ed. A.L. Brown (1987) PHILOCTETES ed. R.G. Ussher (1990) SUETONIUS LIVES OF GALBA, OTHO & VITELLIUS ed. D.C.A. Shotter (1994) TACITUS ANNALS IV ed. D.C.A. Shotter (1989) TERENCE THE BROTHERS ed. A.S. Gratwick (1987) THE SELF-TORMENTOR ed. A.J. Brothers (1988) THE MOTHER-IN-LAW ed. S. Ireland (7990)

THUCYDIDES HISTORY Book II ed. P.J. Rhodes (1988) HISTORY Book III ed. P.J. Rhodes (1994) PYLOS 425 BC: Book IV, 2-41, ed. J. Wilson (1979) WILLIAM OF NEWBURGH THE HISTORY OF ENGLISH AFFAIRS I ed. P.G. Walsh & M. Kennedy (1988) XENOPHON HELLENIKA I-II.3.10 ed. Peter Krentz (1989) HELLENIKA 11.3-IV.2.8 ed. Peter Krentz (1995)