Aristotle (384 BCE–322 BCE)

views updated

(384 BCE322 BCE)

Aristotle was born in Stagira, a Greek colony in Macedonia. His father was physician to the Macedonian king, and the family had both a tradition of learning and connections to the Macedonian elite. At the age of seventeen Aristotle came to Athens to study in Plato's Academy (he may also have briefly studied rhetoric under Isocrates). The community of the Academy included some people who would stay for a few years to learn some philosophy before pursuing political careers in their native cities, and others for whom philosophy was an end in itself, and who might spend their entire lives in the Academy. Aristotle was one of the latter, and stayed in the Academy for twenty years, until Plato's death in 348, when Plato's nephew Speusippus succeeded him as head of the Academy, while the other most prominent Academics, Aristotle and Xenocrates, went to Assos in Asia Minor. There they seem to have formed a kind of local branch of the Academic community under the patronage of the tyrant Hermias of Atarneus, whose niece (and adopted daughter) Aristotle married.

Aristotle spent thirteen years around the north and east Aegean: in Assos; on Lesbos, where he did biological research; in Macedonia, as tutor to the future Alexander the Great; and in Stagira, where he is said to have given laws when it was rebuilt after the Macedonians burned it. He returned to Athens only in 335 (after the Macedonians had attained supremacy over Greece in 338, and after Alexander had succeeded his father in 336), not to the Academy, where Xenocrates had succeeded Speusippus, but to found his own school in the Lyceum, later called the Peripatetic school. He taught there until, after Alexander's death in 323, the Athenians revolted against Macedonia, and Aristotle was charged with impiety for a poem he had written that was held to have given divine honors to Hermias. He left Athens for family property in Stagira's mother-city, Chalcis on Euboia, where he died the following year.

With Aristotle, much more than with Plato, most of the preserved writings are closely connected with his teaching activity. Many of Aristotle's writings bear titles which remain the names of disciplines today (Physics, Politics, etc.), and much of Aristotle's work was either to introduce these disciplines into the Academy and its daughter communities, or to turn them from less systematic practices into systematically teachable disciplines. "Philosophy" in fifthfourth century Athens meant simply "higher education," that is, whatever disciplines, beyond elementary education in gymnastics and "grammar" and "music" (including poetry), might be needed for someone who wishes to live well and to rule his city (or even his own household) well. For different teachers, this would cover different disciplines. For Isocrates, "philosophy" meant rhetoric. For Plato, to judge from the ideal curriculum of Republic VII, it meant mathematics (arithmetic, plane and solid geometry, astronomy, and "harmonics" or music theory) and dialectic (an art of regimented discussion, in which a respondent defends some thesis, typically a definition, and a questioner tries to refute it by yes-no questions leading to a contradiction); these are the means that will lead to knowledge of what really and eternally is, and ultimately of divine things (the Forms and the Good).

Plato conspicuously leaves out rhetoric, which deals with mere opinions rather than with how things really are. He also leaves out pre-Socraticstyle "physics" or "natural history," which he thinks is approximate and probable rather than precise and certain, and which explains things by placing them in a grand cosmogonic narrative of how things come to be, rather than (like mathematics and dialectic) by defining and demonstrating what things eternally are. Aristotle teaches all of these disciplines, without claiming that they are all equally scientific; he introduces a hierarchy of disciplines, from those accessible even to an aspiring politician with no great patience for philosophy, up through more strictly scientific disciplines, to the most demanding but most intrinsically rewarding philosophical wisdom.

Aristotle's introduction of rhetoric (probably already in the Academy) should be seen in the context of the conflict between the Academy and the school of Isocrates. Plato draws a sharp contrast between dialectic and rhetoric: that is, between using question and answer to refute a single respondent on a universal question and using long speeches to persuade a group about such particular questions as are discussed in meetings of a citizen assembly (deliberative rhetoric) or a jury (forensic rhetoric). Plato thinks that only dialectic is worthy of the philosopher. But rhetoric is the path to political success, and so students flock to Isocrates instead. Aristotle thinks that, however narrowly practical many students are, "we Academics," with our philosophical knowledge, ought to be able to educate them better than Isocrates can. (Aristotle is said to have justified his teaching of rhetoric by saying "it were shameful to keep silence and let Isocrates speak," varying a line of tragedy, "it were shameful to keep silence and let barbarians speak.")

This might be merely a practical compromise. More shocking to a Platonist is Aristotle's claim that "rhetoric is the counterpart of dialectic"they are both, not sciences of any one subject matter, but sub-scientific abilities to discover and arrange and express arguments, applicable equally to any subject. Rhetoric also requires rudimentary knowledge of ethics and politics, because these are the subjects about which we must persuade, and because we must know how to project a given character or emotional state, and how an audience of given character and political background will react; the focus remains on argument.

Plato thinks that dialectic, by allowing us to arrive at definitions, gives us a scientific knowledge of eternal Forms existing apart from the sensible world. Aristotle, who has participated in the same dialectical practice as Plato, thinks this claim about its status is spurious. Dialectic is not scientific knowledge of eternal separate Forms, since there are no such Forms. Aristotle is willing to speak of forms present within sensible things (a form is whatever is the object of scientific definition), but dialectic is not scientific knowledge of these forms either, since scientific definition of (say) lunar eclipse depends on specific knowledge of the cause of lunar eclipse, which the dialectician, as a generalist, does not have; it is not the dialectician but the physicist who grasps forms of physical things. Dialectic remains a valuable preliminary training because, by showing what can be refuted, it rules out wrong definitions and helps us find the right ones, and because, by allowing us to find arguments on both sides, it sets out puzzles that science must solve, but it is not itself science or philosophy. And while Plato speaks not of teaching in dialectic, but only of a communal practice of questioning and answering, Aristotle demystifies the practice, and claims in his Topics to teach rules for discovering dialectical arguments, just as his Rhetoric teaches rules for discovering rhetorical arguments.

The average practically minded student will probably study only rhetoric and not dialectic, but Aristotle hopes to lure the better students on further to more scientific disciplines. In the first place, this means ethics and politics, which are philosophical, that is, scientific or causal discussions of what is good for individuals and cities, based on an understanding of what human beings and cities are. But Aristotle distinguishes these "practical sciences" from the "theoretical sciences," that is, kinds of knowledge valued purely for the sake of knowing them, which are capable of greater precision and are more intrinsically worth knowing, though less useful. Against Plato, physics or natural science (in the broadest sense, including biology and psychology) is a theoretical science: when done correctly, it grasps forms of natural things, and proceeds by definition and demonstration, but the forms it grasps are inseparable from matter and motion, and many of its results hold only "for the most part," or ceteris paribus, rather than universally.

Aristotle agrees with Plato that the highest wisdom, the knowledge most intrinsically worth knowing, must be a science of things existing eternally apart from matter, and ultimately of the Good. But neither dialectic nor physics is such a wisdom (nor is mathematics, which is not about separately existing objects, but about ordinary objects hypothetically idealized), and so Aristotle announces, beyond dialectic and physics, a new discipline of "first philosophy" (what commentators of Aristotle since antiquity have called "metaphysics"), which will provide the theoretical wisdom that he thinks both Plato and the pre-Socratics have failed to deliver.


We can broadly divide Aristotle's writings into three classes:

"Exoteric," or "published" writings, were intended for circulation outside the circle of philosophers, elegantly written and sometimes in dialogue form (also the poem for Hermias and a similar poem for Plato). All such writings are lost, but there are substantial fragments; we have perhaps as much as half of Aristotle's Protrepticus, or Exhortation to Philosophy, addressed to a royal patron, which remains an excellent introduction to Aristotelian philosophy. (Aristotle's will is also preserved, in Diogenes Laertius.)

Collections of data, classified but not written up with any literary pretensions, were intended as raw material to be further used in philosophical research and writing and teaching. These texts may have been "loose-leaf," with new material constantly added, some of it perhaps by members of the school other than Aristotle. Extant writings of this type are the History of Animals, the Constitution of Athens (discovered in a papyrus in 1890 and not quite complete, a fragment of a vast series of 158 Constitutions of different cities), and the Physical Problems.

"Acroamatic" writings, that is, writings related to Aristotle's lectures, form the bulk of the surviving corpus. This does not mean that the texts are verbatim identical with the lectures; while Aristotle sometimes speaks as if addressing a live audience, that is compatible with the texts being notes written beforehand as a basis for lectures, or a later revision retaining the lecture style (as in published Gifford or Sather lectures), and the treatises contain many passages which no student then or now could endure if read verbatim as a lecture. The problem is not special to Aristotle; most Greek literature was intended for oral performance, and in every case it is difficult to determine how close the transmitted text is to any given performance. Performances would vary, and the written text is not a transcript of any one occasion but a model for varying expanded or abridged oral performances. In Aristotle's case, while usually only one written version survives for each lecture series, occasionally (as in the Ethics ) we can compare two and gain a sense of the range of variation.

The transmitted texts of the acroamatic writings vary greatly in style. Some passages are highly literary (often marked by avoidance of "hiatus"the juxtaposition of a vowel at the end of a word with a vowel at the beginning of the nextas in Isocrates and in Plato's late dialogues), whether because they have been more thoroughly revised toward eventual publication, or because Aristotle delivered some pieces (especially the beginnings of works) in more elaborate form, or because they are excerpted from Aristotle's exoteric works. Other passages are long strings of brutally truncated arguments for the same conclusion, connected merely by "also"; in performance Aristotle would have selected only some arguments, and filled them out and connected them better.

The transmitted texts contain many references to "what we have said previously/elsewhere" or "what we will say," sometimes with a title "in the [writings or lectures] on x." (It is possible, but should not be the default assumption, that some of these cross-references were added by later editors.) While we can often supply a plausible page reference, we should beware of assuming that Aristotle's references are to texts now extant, or else to lost parallel texts: They are not necessarily to fixed texts at all, but to earlier and later parts of an idealized curriculum, each part of which would be repeatedly given (with variations) as a lecture, and also written down and occasionally updated, even if no actual student ever heard the whole series in order. "We have said" and "we will say" refer not to order of composition but to order in the curriculum; however, while Aristotle is mostly consistent about the ideal order, there are contradictions that may indicate that he changed his mind on the appropriate sequence of the psychological-zoological writings. There is no real contradiction in the fact that Aristotle (and his followers) cite the same work under different titles; the curriculum may be subdivided more or less finely, and the same title may be used generically for a large section or specifically for a smaller subsection: "physics" or "on nature" may refer to the entire physical-biological corpus or to something as narrow as Physics IIV (with Physics VVIII cited contrastively as "on motion").

Some ancient catalogs list Aristotle's works by shorter units and some by longer units (the catalogs may also contain duplications, and some catalogs refer to works not available to other catalogs, or to us). Andronicus of Rhodes in the first century BCE attempted to introduce order by determining the correct titles and sequence, generally opting for longer "works," and it is probably roughly his decisions that won out. (But the story that Andronicus, drawing on a rediscovery of Aristotle's library, made the acroamatic works available for the first time and so touched off a renaissance of Aristotelianism, is mostly or wholly fiction.) Following a Stoic division of philosophy, Andronicus organized the corpus into first "logical" writings, then "physical" (or more broadly "theoretical" writings, to include the Metaphysics, concerned with nonphysical things), then "ethical" writings (or more broadly "practical," to include the Politics ).

Many of the texts are now lost. As with the rest of Greek literature, what survived was generally only what was used and copied for educational purposes, which explains why the "exoteric" works are lost and why usually only one version of each "acroamatic" text survives. The surviving texts have been edited many times since the invention of printing, often in complete editions that generally try to follow Aristotle's and other ancient indications of the correct sequence of the corpus (although these are not fully consistent and, for example, give no hint how to order the three surviving ethical works, which all fill the same place in the curriculum).

Immanuel Bekker's nineteenth-century edition has become standard. Modern editions and translations give "Bekker pages" in the margins (e.g., "1042b5," where "a" or "b" is a column of a double-columned page), and Aristotle is always cited in this form where possible; editions that aspire to completeness print the texts in Bekker's sequence. The editions divide Aristotle's treatises into books and chapters; the book divisions have (not always undisputed) ancient authority and may in some cases go back to Aristotle himself, but the chapter divisions are modern artifacts and deserve no deference (medieval authors use a different division into "lectiones"). Ancient writers cite the books of multibook treatises by Greek letter-names; modern writers generally use numbers, but prefer letters in the Metaphysics, where the presence of two books alpha (conventionally designated Α and α) disrupts the usual letter-number conversion.

The following list presents the texts in Bekker's sequence, leaving out texts currently agreed to be spurious, and marking with an asterisk texts whose authenticity is currently controversial. The traditional Latin titles are added where these sound significantly different from the English.

Logical writings (Organon ): Categories (the title is controversial), On Interpretation, Prior Analytics, Posterior Analytics, Topics, On Sophistical Refutations (De sophisticis elenchis ).

Theoretical writings: Physics, On the Heaven (De caelo ), On Generation and Corruption, Meteorology, On the Soul (De anima ), Parva naturalia (including On Sense and Sensibilia, On Memory, On Sleep [De somno ], On Dreams [De insomniis ], On Divination in Sleep, On Length and Shortness of Life, and On Youth, Old Age, Life, Death and Respiration [De juventute for short]), History of Animals, Parts of Animals, Movement of Animals, Progression of Animals (De incessu animalium ), Generation of Animals, *Physical Problems, Metaphysics.

Practical writings: Nicomachean Ethics (abbreviated "NE" or "EN"), *Magna Moralia, Eudemian Ethics ("EE"), Politics. (In a peculiar situation, three central books are identical: NE VVII = EE IVVI. These books are usually printed with the NE, but most modern scholars agree that they were originally written with the EE instead.)

Bekker puts at the end the Rhetoric and Poetics, under the head of "productive philosophy" (i.e., philosophy to guide production, in this case of speeches or poems); their place is controversial, and they had sometimes been put at the end of the Organon. The *Constitution of Athens and other texts not printed by Bekker (fragments discovered on papyrus or in later ancient citations or translations) are often placed at the end.

With Aristotle, as with Plato, there have been attempts to determine the order of composition of the works, distinguishing "early," "middle," and "late." Sometimes stylometric tests are applied. Some scholars, like Jaeger, assume that Aristotle moved from an early Platonism, to a critical revision of Platonism, to an independent mature philosophy. Such "developmental" studies have had the merit of bringing out tensions in Aristotle's work, and calling attention to works (often fragmentary, like the Protrepticus ) that had been ignored or deemed spurious because they seemed embarrassingly close to Plato. Some chronological results have won widespread assent, notably that the Protrepticus is early, and the EE earlier than the NE. But dating has not generally been successful, and for a good reason, namely that Aristotle regularly revised his work, so that a single text may show both "early" and "late" features and thus resist easy classification. Aristotle was trying to present his treatises as parts of a synchronic system, ordered by pedagogical role; tensions remain, and while sometimes these tensions are best explained diachronically, this is not always the case. In what follows, the most important texts will be discussed, not in Bekker's order or in a presumed chronological order, but in roughly the order of increasing difficulty: probably many of Aristotle's students dropped out early in this sequence, and only a few remained until the end.

Ethics and Politics

Aristotle conceives ethics as a part of political philosophy: We cannot understand and evaluate different political structures unless we understand individual character, and conversely, we cannot fully describe the best life for an individual without reference to the city in which he lives and is educated. Many comments in the ethical works assume that the reader or hearer is (or wants to be) a politikos, or statesman, and Aristotle assumes that the best life for an individual and the best politeia or constitution for a city, whatever they turn out to be, will be analogous. The ethical works, then, emerge from popular lectures to aspiring politikoi, who have come to hear lectures by a philosopher in the hope that it will make them happier and more successful politikoi, but who do not intend to spend their lives on philosophy.

Aristotle can be seen as trying to repair the damage that Plato did in his lecture on the Good, where an audience who had come expecting to hear about "health or wealth or some marvelous happiness" were surprised to find that the lecture was about numbers and that its conclusion was that the Good was the One, with the result that some of the audience gave up on philosophy altogether, while others presumably turned to the more practical philosophy of teachers like Isocrates (see Aristoxenus, Elements of Harmonics II, 1, and cf. EE I, 8). Aristotle is in part rejecting Plato's conclusions (he thinks mathematics has nothing to do with goodness), in part simply rejecting his method of presentation: we must start with what the audience antecedently believes and values, get them to see the difficulties, and so introduce philosophical doctrines (including any doctrine of a higher good) as solutions to those difficulties. But in ethics, as in rhetoric, he thinks that the Academics should be able to educate them better than Isocrates can.

Anyone who can choose how to live, and who wants to approach the question rationally, must first clarify what he is aiming atwhat is the chief good of human life. Everyone agrees that the aim is eudaimonia usually translated as "happiness," but perhaps best neutrally as "success"; it need not be introspectible, must be evaluated over a lifetime rather than at one moment, and can be said of cities as well as of individualsbut they disagree about what eudaimonia consists in. The three plausible contenders for the best way of lifethe pleasure-seeking life, the active or political life, and the contemplative or philosophical life (Aristotle thinks the money-making life is chosen only from necessity)go with different contenders for the human good. The pleasure seekers think it is pleasure; the politikoi may think it is fame or honor or, more appropriately, that it is aretê, virtue, or excellence (what deserves honor).

Among the philosophers, Socrates thinks that virtue (consisting in some kind of knowledge) is necessary and sufficient for happiness, and Plato talks about the Form of the Good or about the One. Aristotle creates an aporia by using these views against each other and raising objections against each, in order to motivate his own account of happiness, and the conceptual distinction on which it is based, as a solution to the aporia. Happiness or success in life is not virtue, which is a stable hexis ("habit" or acquired state) persisting even when it is not exercised, but rather the energeia (exercise or activity) of virtue throughout a complete lifespan. We can thus avoid the paradox of saying that the good person is happy even when poor, sick, and unjustly despised by his fellow citizens; in such a condition he remains virtuous but cannot exercise his virtue, or is greatly hampered in exercising it. The happy life will involve virtue, and it will also involve wealth if the virtues (say, generosity) need wealth to be exercised, and these facts explain the temptation to identify happiness with virtue or even wealth. Likewise, the happy life is pleasant, since Aristotle analyzes pleasure as being (or following upon) the exercise of a natural state, but its pleasantness is not what makes it happy or worthy of choice. (This is against the view of some Academics that pleasure is always a process, the restoration of a natural state, and that the happiest life is a steady natural state without deficiency or restoration. Aristotle avoids the paradox that the happiest life is without activity or pleasure by arguing that there are energeiai that are not processes.)

Aristotle applies the same method of setting out competing beliefs and arguments, resolving the aporia through a distinction, and showing how justice can be done to all sides, to resolve Socrates' paradoxical argument that incontinence is impossible: I can do something wrong if I have hexis -knowledge that this type of action is wrong, but not if I am applying the hexis and have energeia -knowledge that this particular action is wrong. It must be stressed that this is a teaching method, designed to motivate Aristotle's doctrines and conceptual distinctions for his audience and to make softened versions of Socratic or Platonic paradoxes more palatable. We do not know that this is how Aristotle himself arrived at his conclusions.

Aristotle also tries to show what is right in the Socratic and Platonic conclusions that virtue and happiness consist in knowledge, perhaps knowledge of a transcendent Good. The work or task or function (ergon ) of a human being is rational activity, and a virtue is a condition that disposes to such activity. But there are two kinds of virtues: "intellectual virtues," or virtues of the rational soul, and "moral virtues," conditions of an irrational part of the soul, according to which it is disposed to act as reason would require.

Genuine moral virtue is not simply habituation to desire the right amount, but involves choice, which involves deliberation or means-end reasoning: so moral virtue is not possible without the intellectual virtue of phronêsis ("prudence," "practical wisdom") or deliberative ability. (Nor, conversely, is phronêsis possible without moral virtue, since uncontrolled passions will warp our deliberations.) But phronêsis is not identical with the highest intellectual virtue, sophia ("wisdom"), knowledge of the divine things that are intrinsically most worth knowing. Sophia is exercised only in contemplation (theôria ) and not in action: we cannot deduce, from these necessary eternal things, knowledge of the contingent temporal objects of practical choice. But sophia gives a starting point for deliberative reasoning in another way, because contemplation is itself the exercise of the highest virtue, and is therefore the highest happiness we can try to achieve. (It is the only exercise of virtue we can attribute to the gods, who can hardly be courageous or temperate.) So while happiness is possible with only moral virtue and practical intelligence, the highest happiness needs theoretical intelligence as well.

When Aristotle says that maximizing contemplation is the highest goal of human planning, he means not only planning an individual life, but also a statesman's planning for the city. (The statesman may have only phronêsis, but needs proper respect for sophia.) Happiness, for cities as for individuals, is an exercise of virtue, and while this may require material conditions (prosperity and external peace), the statesman's main concern should be making the city virtuous. And, for cities as for individuals, some virtues are more worth exercising than others: courage and military solidarity are virtues we would rather not have occasion to exercise. While a city must be able to defend itself, its highest goal is the exercise of the virtues of peaceful leisure. This is theôria, not only in its metaphorical sense (the philosopher's contemplation of nature or of incorporeal divine things), but also in its ordinary sense: attendance at civic religious festivals, including the musical-poetic contests (of tragedies, comedies, etc.), which may be occasions for private or communal moral and political reflection. (The Poetics defends the value of such musical-poetic performances, and inquires how it comes about; it thus elaborates an important point too briefly treated in the Politics.)

Aristotle's main goal in the Politics is the construction of an ideal politeia (constitution or collective mode of life and governance), a critical revision of Plato's Republic and Laws. But he also discusses less ideal politeiai, how they are preserved by proper legislation, and how they are corrupted, leading to revolution; the trained politikos will be useful even to a non-ideal politeia, helping to preserve it by moderating and improving it. (And Aristotle's 158 collected Politeiai will help give an empirical base.) The central thesis of the Politics is the distinction between genuinely political rule (rule over free fellow citizens, in the interest of the ruled) and despotic rule (rule as of a master over slaves, in the interest of the ruler). While Politics I is notorious for defending slavery, Aristotle's main interest is to make clear the differences between despotic rule (legitimate only within the household) and political rule. (He thus also defuses the Socratic paradox that there is only one art of ruling, depending on philosophical knowledge of the good: to the Athenian bourgeois, this suggests that the Academics are claiming the right to rule over their fellow citizens, while not allowing ordinary citizens even to give orders to their servants.)

Within the city, not only tyranny but also oligarchy and democracy are despotic: even when they are ruled by law, their laws express the economic and political interests of a ruling individual or group (the rich few in an oligarchy, the poor majority in a democracy). But rather than conclude, with Thrasymachus, that all rule is despotic, Aristotle argues, with Plato, that genuinely political rule is possible. Officially (like Plato's Statesman ) Aristotle has a two-by-three grid of constitutions: corresponding to tyranny, oligarchy, and democracy are three good constitutions, kingship, aristocracy, and (what Aristotle calls in a narrower sense) politeia, which are the rule of the one, the few, or the many in the interest of the whole city. But actually Aristotle treats kingship and aristocracy as an ideal constitution run by morally and practically virtuous people and aiming at the development and exercise of virtue; politeia is a more attainable ideal, a "mixed constitution" between democracy and oligarchy, as the moral virtues lie between vices of excess and deficiency. Politeia, though a "virtuous" constitution, does not aim at virtue in the citizens and does not choose officials for their virtue, but at least its laws, balancing the interests of different groups and designed to preserve peace between them, do not impose a partisan "justice" that would conflict with genuine moral virtue in the individual.

Dialectic and Analytics

Aristotle's logical treatises are usually grouped as the Organon, or "instrument"; against the Stoics, who make logic a part of philosophy alongside physics and ethics, the Peripatetics say that logic is a mere instrument of philosophy, valuable neither intrinsically nor as guiding action, but only as guiding reasoning in other fields. Also ancient is the arrangement of the Organon : first the Categories, dealing with single terms and the simple objects (substances, quantities, qualities, relations, actions, passions, "where," "when," positions [e.g., standing], states [e.g., armed]) that they signify; then the De interpretatione, dealing with propositions composed of two terms linked by a copula (affirmative or negative, universal or particular, assertoric or modal); then the Prior Analytics, dealing with syllogisms, valid arguments composed of three propositions sharing three terms (e.g., "A belongs to no B, C belongs to all B, therefore A does not belong to all C," valid since Aristotle rejects empty terms).

Then come treatises dealing with different types of syllogism: the Posterior Analytics, with scientific or demonstrative syllogism, where the premises must be true and causally explanatory of the conclusion; the Topics, with dialectical syllogism, where the premises need only be plausible; the Sophistical Refutations, with sophistical or pseudo-dialectical syllogisms, which are only apparently valid or have only apparently plausible premises; some ancient writers add "rhetorical" and even "poetic" syllogisms. At the end of the Sophistical Refutations, Aristotle says that while he has perfected earlier teaching of rhetoric, in the case of the syllogism there had been no such teaching before him; Aristotle has been taken as here summing up the Organon and reflecting on his crucial discovery, the syllogism.

However, Aristotle has no conception of "logic," but of two different disciplines, analytics (Prior and Posterior ), and dialectic (the Topics, taken as including the Sophistical Refutations ); the Categories and De Interpretatione seem designed to support the Topics rather than the Analytics. We have spoken above of dialectic, the practice of regimented discussion in which a questioner seeks to refute a respondent's thesis by a series of yes-no questions. The end of the Sophistical Refutations is summing up not the entire Organon but only the Topics, which has for the first time made dialectic a teachable art and has shown how to discover syllogisms to deduce the contradictory of the respondent's thesis.

These arguments, unlike rhetorical arguments, can proceed only from premises the respondent will grant, and by steps he must accept as valid. Dialectic must proceed from plausible (endoxa ) premises, since these are just those premises that a respondent will concede (if he does not see that they favor or hurt his thesis). It is a mistake to turn dialectic into "argument from prereflective intuitions," detached from the context of refutation, and to give it a foundational role in philosophy. Aristotle does say that dialectic gives a path to the principles of the sciences, but these principles are, especially, definitions, and, as in Socratic dialogue, dialectic is chiefly devoted to testing and refuting proposed definitions. The structure of the Topics brings this out: successive books give rules for testing claims that P belongs to S, that P is or contains the genus of S, and that P is proper (idion ) to S (i.e., belongs to every S and no non-S ), which are necessary but insufficient conditions for P to be the definition of S, and then give special rules for testing claims of definition. Aristotle also gives advice on how to order your questions, how to proceed as respondent, and background knowledge the dialectician should have.

The Categories and De interpretatione, as well as Topics I, seem to give such background knowledge; the most recent edition of the Categories prefers the alternative ancient title The Before the Topics, in part because the text is not just about categories. First, Aristotle distinguishes simple from complex expressions; then, what is signified by a simple expression is signified either synonymously (univocally) or homonymously (equivocally) or paronymously (denominatively). Two things are synonymous if they are signified by the same name and according to the same definition; homonymous if signified by the same name according to two different definitions (bank and bank, but also mousikê, the art of music, and mousikê, a female musician); paronymous if one name is derived from the other ("just" is paronymous or derived from "justice," not because the word "justice" is older, but because something is called "just" because there is justice in it).

Only synonymous things, not homonymous or paronymous, can be given genus-differentia definitions ("just" is neither a species of animal nor a species of virtue). Synonymous things that are in a subject (like justice) fall under one of the nine categories of accidents; synonymous things that are not in a subject are substances. (Substances can be "said of" something, but cannot be "in" something: horse is said of Bucephalus, since Bucephalus is said to be a horse, but there is not a horse in Bucephalus. "Primary substances," like Bucephalus, are neither in anything nor said of anything.) Aristotle gives tests for when a thing falls under each category, which are needed to apply the rules of the Topics (thus if P is the genus of S, S and P must belong to the same category, but we need tests to determine to which category they belong). Likewise, after the categories proper, Aristotle gives accounts of the different kinds of opposition, priority and simultaneity, motion and having, which serve similar functions in dialectic.

Sometimes a dialectical questioner poses a series of questions that appear to necessitate the contradictory of the respondent's thesis, but which contain some hidden fallacy; the respondent must avoid assenting to what does not follow, and must be able to explain why it does not follow, in order to avoid appearing, to the spectators and perhaps even to himself, to have been refuted. The Sophistical Refutations, which may be considered as a final book of the Topics, is devoted to classifying such "sophisms," or "sophistical refutations," explaining how each type arises, and advising the respondent on how to recognize and to "solve" or "resolve" each such sophism as it comes at him in questioning.

Sophisms are not intrinsically dishonest: They are puzzles demanding solution. We should imagine, not an arms race between sophists devising offensive weapons and philosophers improving defenses, but a single intellectual community exploring sophisms and discussing the merits of different possible solutions. Often the most philosophically interesting sophisms are "sophisms of figure of speech," arising when the grammatical form of a term misrepresents its logical form: these include the family of sophisms concluding that "there is a third man" beyond mortal individuals and the Platonic Form, which turn on treating "man" as "signifying some this." Aristotle himself, in the fragmentary On Ideas, constructs a series of such philosophically serious sophisms, giving for each Platonic argument for the Forms a parallel argument to an unacceptable conclusion, such as the third man. Each sophism challenges the Platonists: "dismantle my sophistical argument without at the same time dismantling your own allegedly probative arguments for the Forms." The Categories helps solve sophisms of figure of speech by testing what category each term signifies, and its distinction between primary and secondary substances can solve many third man sophisms; but if Platonists accept these solutions, they risk undermining their own favorite arguments and conclusions.

A syllogism or deduction is "a discourse in which, some things being supposed, something different results of necessity through their being so." Syllogisms are as old as thought and language, and Aristotle does not claim to have invented them. What the Analytics invents is a method for analyzing them: that is, for classifying them and then, by giving a few primitive argument forms and derivation rules for generating more complicated forms, explaining why syllogism comes about. In every case, syllogism depends on two premises sharing a common term (the syllogism will be in different "figures," depending on whether the shared term is subject of one premise and predicate of the other, predicate of both, or subject of both; some "moods" will be valid and others not, depending on whether the premises are affirmative or negative, particular or universal, assertoric or modal). Aristotle's analysis depends on the realization that the necessity or validity of an argument, once all premises are made explicit, depends only on its form, so that the same analysis applies whether the premises are true or false; this realization presumably arose from the deliberate exploration, in dialectic, of the consequences of false hypotheses.

But Aristotle sharply distinguishes dialectical from scientific or causal reasoning, and he devotes the Posterior Analytics to analyzing "demonstrations" or scientific syllogisms, arguments that give their possessors knowledge or science (epistêmê ) of some object; here epistêmê is a cognitive state that not only grasps an object as it is, without the possibility of falsehood, but also understands why the object is as it is. It seems surprising that mere arguments, without direct contact with the object, can give such knowledge, and Aristotle tries to analyze the conditions under which this can happen. The premises must be true, necessary, and better known than the conclusion; they must also express the causes that explain why the conclusion is true. We can of course come to know an object by reasoning from effects to causes, but properly scientific and explanatory knowledge must reason from causes to effects; the logical structure of the argument will mirror the causal structure of the world.

On pain of circularity or regress, the first principles of demonstrations must be known by some means other than demonstration (Aristotle calls the non-demonstrative grasp of first principles nous rather than epistêmê ). Apart from some topic-neutral principles of reasoning ("axioms"), these will be either "hypotheses" that the objects of each science exist, or "definitions" of those objects; we accept without demonstration both the existence and definitions of the simple objects of the science (e.g., for geometry, point and straight line) and preliminary definitions of the complex objects (e.g., regular pentagon), but we demonstrate the existence of complex objects satisfying those definitions. Dialectic can reach these preliminary definitions, but we can give properly scientific definitions of complex objects only once we demonstrate their existence from simple causes (thus not "thunder is noise in the clouds" but "thunder is noise of extinction of fire in the clouds"). We cannot give justificatory explanations of how we know the first principles of the sciences, but only causal explanations of how the human mind, primed by experience, comes to grasp them by nous ; Aristotle's account is compressed enough that it has been read both as an empiricist account of induction and as a friendly revision of Plato's theory of recollection.

Aristotle's account of science is clearly modeled on geometry. But he tries to show that physics too can be a science, beginning from a grasp of the forms of natural things.

Physics and Cosmology

Aristotle's project in physics is a response to Platonic challenges both to the narrative method and to the content of pre-Socratic physics. Anaxagoras's physicsto take a typical pre-Socratic examplenarrates the origin of everything from a cosmogonic vortex, whose rotation and centrifugal force explain the separation of heaven from earth, the rotation of the heavens, the motion of heavy bodies down and light bodies up and the sorting of like bodies to like, and then the formation of the first plants and animals and humans out of seeds present in the precosmic mixture. Plato thinks such narrative can never be scientific; science must be concerned not with how things come to be but with what they are, beginning from their forms as grasped by definitions, and proceeding to demonstration.

Plato also complains that pre-Socratics explain the emergence of the cosmos by reference not to a rational plan or to some good to be accomplished, but through violence; if things are where they are because of a vortex (i.e., through being shoved by other bodies that are shoved into them) rather than because it is best for them to be there, then there will be no explanation of the goodness and orderliness of the universe, as manifested in the mathematically precise motions of the planets. In the Timaeus, Plato addresses the second objection by sketching an alternative teleological physics; but this too follows a narrative method, and even a reformed physics cannot be science but only a likely story.

Aristotle tries to address both objections and to produce a genuinely scientific physics, explaining the physicists' traditional explananda (rotating heavens, fall of heavy bodies, lightning, earthquakes, animals ) not in a narrative sequence but in a causal or explanatory sequence, beginning from the form or nature of each body, which is the object of a properly physical definition. Aristotle broadly accepts the Timaeus 's picture of the cosmos: a spherical earth is at rest at the center of a single spherical cosmos. The cosmos is made of earth, water, air, and fire intertransformed and combined, teleologically organized to support living things, and surrounded by heavenly bodies that are themselves living and divine; these move in several uniform circular motions, which combine to produce complex astronomical phenomena, and they are ultimately governed by an incorporeal god or gods. But Aristotle's method contrasts with the Timaeus, and leads him to challenge particular claims of the Timaeus as well as of the pre-Socratics.

Aristotle's particular physical treatisesthe De caelo, On Generation and Corruption, Meteorology, and psychological-zoological writingsfollow roughly what had been the traditional narrative sequence of the explananda. Thus the De caelo treats the rotation of the heavens and the motions of heavy and light bodies, traditionally explained through a cosmogonic vortex. But Aristotle rejects explanations through vortices or any other violent cause. What happens to a thing violently, contrary to its own nature, cannot happen always or for the most part, but only as a temporary interruption of a thing's natural behavior (e.g., a stone being thrown upward). Physics, as a science, seeks to explain what happens always or for the most part, and must therefore start by grasping the nature of each thing, where "nature" means "principle of natural motion"; so the nature of heavy bodies is to move toward the center of the cosmos, and thus teleology is built into each nature. (Thus physical definitions necessarily involve motion, and the forms they describe cannot exist separately from matter, as the Form described by a Platonic dialectical definition is supposed to. And fire and so on must be defined physically by their motions, rather than mathematically by their shapes as in Democritus and the Timaeus.)

Aristotle draws the conclusion that the heavens cannot be made of the four standard elements; since these naturally move in straight lines toward or away from the center, the heavens would have to be constrained to circular motion by violence (whether by a vortex or by a providential soul as in the Timaeus ), and such motion could not be regular or permanent. Consequently, the heavens are made out of a fifth element (sometimes called "aether") whose natural motion is around the center. The aether is free of the accidents that obstruct natural motion in the sublunar world, so it rotates eternally without interruption or irregularity. Because this motion arises eternally from the nature of the thing, Aristotle rejects the claim of the pre-Socratics and the Timaeus that the rotation of the heavens and the separation of the elements into an ordered cosmos arose (by what could only be a violent process) from a precosmic chaos; the ordered world and its more-or-less regular phenomena have always existed, and a narrative explanation is excluded, since there is no precosmic situation from which a narrative could begin. Rather, the phenomena must be explained by the influence of the naturally rotating heavens on naturally moving sublunar elements.

The On Generation and Corruption and Meteorology continue this program. If it were not for the rotation of the heavens, the four sublunar elements would separate out into concentric spheres of heavier and lighter elements, with no intertransformation or combination, and therefore no living things. But the regular daily rotation of the heavens, combined with the regular rotation of the sun through the inclined circle of the ecliptic, bring it about that the sun is above the horizon more of the time in the summer than in the winter, causing regular cycles of heating and cooling, and thus of evaporation and condensation.

Aristotle sees evaporation as a genuine transformation of water into air, and likewise of earth into fire; when a heavy element is transformed into a light element, it begins to rise (and when a light element is condensed, it falls), and this cycle keeps the elements from separating and gives rise to combinations. But properly the light elements are not "air" and "fire" but "moist exhalation" and "dry exhalation"; air is a mixture of both, and the portion of the dry exhalation that gathers above the air and beneath the sphere of the moon is not actually fire, but is a fuel that easily becomes inflamed, as it does in comets and shooting stars.

Since Tycho Brahe proved that comets and novae are supralunar, Aristotle's account has been regarded as a desperate attempt to save his theory of immutable heavens by moving all changes in the heavens to a fictional sublunar fire-sphere governed by a fantastic exhalation process. Historically this is the wrong attitude. Aristotle's explanation of comets is among the most traditional parts of his physics: Heraclitus explains even the sun through a continuous process of exhalations rising from the sea and becoming inflamed. Aristotle's innovation is to separate out from meteorological phenomena genuinely astronomical things like the sun, which are not dependent on the sublunar world but are governed only by themselves and by unchangeable incorporeal things, and therefore have eternally constant motions and can be objects of precise mathematical science; it is only because these things are perfectly regular that they can impose even an approximate regularity on the sublunar world.

The Physics in the narrower sense is a deliberately non-cosmological prolegomenon to the physical works, describing the principles from which all natural things arise and the necessary conditions (above all, motion) for anything to arise from these principles, and using a definition of "nature" to delimit the physicist's domain and methods and the causes or explanations that he must invoke in tracing natural things back to their principles. Aristotle begins, traditionally enough, with the archai, the principles or starting points of natural thingswhatever must exist before each natural thing comes to be, and can be used in explaining it. (For narrative physics these would be whatever existed before the cosmos, e.g., for Empedocles the four elements and love and strife, for the Timaeus the Forms and receptacle and demiurge; but Aristotle's archai do not exist before the cosmos, since his cosmos never came to be.) We will infer to the archai by analyzing the characteristic effect that arises from them, which is, most generally, motion or changenot only change of place (locomotion) but also change of quality (alteration), change of quantity (growth and diminution), and the coming to be and passing away of substances (generation and corruption).

Aristotle argues that whenever some new F comes to be, in any category, there must be some persisting substratum that was not F and comes to be F ; this analysis shifts F to predicate position. The subject that persists through even substantial change is one archê, the matter. This echoes the Timaeus 's argument that the apparent substantial change of (say) water into air shows that the real archê is not water or air, but the receptacle, the persisting substratum that appears now watery, now airy. But the Timaeus seems to infer that the change is not really substantial, that all sensible things are just accidental modifications of this single persisting substance. Aristotle argues that there is real substantial change, that the substance of a natural thing is not the matter that persists through the thing's generation and corruption, but the form that comes to be in the matter. Both form and matter are archai of natural things, and while the matter is potentially this or that substance, the form, as what makes each substance actually that substance, is substance in a stronger sense. (Plato would reply that while form as well as matter is an archê and a substance, the real Form is eternal and separate, and what comes to be in the matter is a nonsubstantial image of the Form.)

How do we tell when a form acquired through change is a new substance, and when it is merely a new accident of a persisting substance? The shape of an artifact is merely an accident, but the nature of a natural thing, that is, the distinctive "principle of motion and rest" within it that is responsible for its carrying out its characteristic activities, is a substance. Physics II argues that the nature of a natural thing is more properly its form than its matter, and therefore that the physicist must study form as well as matter; thus, as we have seen, physics must define and not merely narrate, giving definitions that, unlike Platonic dialectical definitions, are inseparable from motion and thus from matternatures are like "snubness," which is neither the matter "nose" nor the form "concave," but a form that cannot be defined without reference to its appropriate matter, the nose. A natural thing acts for the sake of actualizing the characteristic potentialities of its nature, and so the physicist will give explanations not only through the material and formal causes and through the mover or efficient cause, but also through the final cause. Aristotle thus, like Plato in Laws X, argues against many pre-Socratic physicists that purposive activity is prior to chance and violence, but he does this while preserving what is specific to nature, and without reducing natural things to artifacts of a designing soul.

Nothing will arise from matter and form without motion; motion depends on time and place and (some people think) on void; also a motion, to be a single motion, must be continuous, and continuity implies infinite divisibility. All these concepts are problematic, and Aristotle tries to define, and to resolve aporiai about, motion, place, and time, and to show that the infinite and the void do not exist (except in specially qualified senses). He then turns to the "On Motion," Physics VVIII (Physics VII seems to interrupt the argument, and may be a survivor of an earlier stage of Aristotle's work). Physics VVI give non-causal considerations that would apply equally to natural and violent motion, notably about when a motion is a single motion, about when two motions or a motion and a rest are contrary, and about the continuity of motion, place, and time; they seem to be there chiefly to supply premises for the causal argument of Physics VIII. Physics VIII, relying only on the abstract concepts of the Physics and not on empirical cosmology, gives an elaborate argument from the natural motions of corruptible things, first to the eternity of motion as such, then to self-moved movers (empirically, animals) and unmoved movers (their souls), then to an eternally continuous motion (the motion of the heavens), and finally to an eternally unmoved cause outside the cosmos. This bravura display reaches beyond physics to metaphysics or theology, and Aristotle relies crucially on it in Metaphysics Λ, discussed below.

Psychology and Zoology

Narrative physics typically ends with the production of plants and of animals, including humans, before turning to human societies and conventions, which Aristotle treats under practical philosophy. Aristotle devotes a large part of his writing to animals, complemented by Theophrastus's studies of plants. But his program of denarrativizing physics, and of physical teleology and physical definition, entail major differences from earlier accounts of animals; Aristotle also integrates an account of soul into his study of animals, though not as fully as we might expect. The crucial methodological texts are Parts of Animals Book I, which serves as an introduction to the zoological works generally, and De anima I, 1.

A narrative physicist believes he has accounted for the elephant once he has taken the cosmogonic narrative far enough to generate the first elephant. This means that he puts his "Generation of Animals" before his "Parts of Animals." (The parts of an elephant are simply whatever results from the prior generative process: Thus the Hippocratic On Fleshes gives a cosmogonic account of the generation of each tissue, with no regard to how the tissues are arranged in the animal, what animal they are parts of, or what functions they have.) Such a physicist will also be more concerned with the hard problem of the "spontaneous" (nonsexual) generation of the first elephant than with the easier problem of how to get more elephants out of the elephants there already are.

For Aristotle, however, the whole cosmos with all its species has existed from eternity, so there is no reason to believe elephants were ever generated spontaneously. We never see elephants generated spontaneously anymore, and while nature might have had greater generative force at some past time when it was undergoing more violent motions (see Physical Problems X, 13), when we understand the extremely complex arrangement of parts required for a functioning, self-sustaining elephant, it becomes incredible that the crude natural powers of the pre-elephantine era could have combined to produce it. (Plato might say that God intervened to produce the first elephant, but Aristotle thinks that God acted no more or less then than now, and that his activity simply sustains the regular activities of natures. While Aristotle is now notorious for defending spontaneous generation, he actually allows less scope to spontaneous generation than any other Greek philosopher, restricting it to lower life-forms.)

Thus when Aristotle studies the generation of living things, he is chiefly studying their generation out of already existing members of the same species. And we can understand this process not in narrative sequence but only backward, starting from the arrangement of parts that the generative process is for the sake of producing; so methodologically the Parts of Animals must precede the Generation of Animals. And the parts themselves must be explained teleologically, not through the generative process but through their function in the animal. Different species of animals will have different strategies for survival and reproduction, thus different characteristic activities, requiring different characteristic parts; the scientist will define each animal species by describing its characteristic parts, defining each part as an "organ" or instrument of some activity and deducing its shape and matter from its function.

Aristotle describes the parts, and the whole animal, as organs of the soul, that is, instruments through which the soul's powers are exercised. Because they cannot be defined without reference to the soul, it belongs to the natural scientist to study soul, or at least those powers of soul that are exercised through bodily instrumentsall powers except, possibly, intellect (nous ). Aristotle is trying here both to reform physics by making it include the soul, and also to make the study of soul scientific by bringing it under physics. However, he also makes the study of soul further from physics as usually conceived, by denying that the soul is moved, either in moving the body or in sensing and thinking. In De anima I he says that earlier philosophers have approached the soul either from its capacity to originate motion in the body, concluding that it is a self-moving source of motion; or from its ability to represent all things, concluding that it is composed of the elementary constituents of all knowable objects; or from its "bodilessness," identifying it either with fire or air or with something entirely incorporeal.

The Timaeus combines all of these approaches but, Aristotle thinks, in a mistaken way, representing the soul as a magical quasi-body interwoven with visible bodies, moved in the same way that bodies are, and moving bodies and being moved by them in the same way that bodies move each other. In De anima II, Aristotle instead defines the soul by its relation to its energeiai, the activities it carries out through the body. Soul is the dunamis (power, potentiality, capacity) for these energeiai, or it is that which, added to a potentially living thing (a seed or embryo), makes it an actually living thing, where to be an actually living thing is to have the potentiality to carry out an appropriate range of the vital activities (nourishment, growth, reproduction, sensation, memory, imagination, desire, locomotion, intellection). In Aristotle's formula, soul is "the first actuality [entelecheia ] of a potentially living body," the second actuality being the vital activities; soul stands to these activities as a hexis of science stands to the exercise of that science in contemplation, or as a productive art stands to its exercise in production.

Aristotle spells out his definition by saying that soul is "the first actuality of a natural organic body." Modern connotations of "organic" are misleading here: an organic body is an instrumental body, as is, for example, a hammer; the living body is the instrument of the soul as the hammer is an instrument of the art of carpentry. But the hammer is an artificial organic body, while the living body is a natural one, meaning (by the definition of the Physics ) that it has an internal principle of motion and rest. So while the art of carpentry moves the hammer from outside (by inhabiting the body of the carpenter), the soul is a nature moving the body in a quasi-artistic way from inside, in producing and maintaining its natural instrument (nutrition, growth, reproduction) and in further using that instrument (sensation and the higher activities). The arts give us a model for how the soul can move its body without itself being moved (unlike a body pushing or pulling another body): though the carpenter's hand is moved when he moves the hammer, his art of carpentry is not. The arts also give a model for the cognitive powers, since an art contains the "formula," the definition or perhaps recipe, of its objects, without containing their matter; and arts can recognize individual objects through cognitive instruments (the art of measuring might use scales), as well as moving them through instruments of action.

The vegetative powers (powers shared even by plants) and the sensitive powers (powers shared by irrational animals) are "not without" their appropriate bodily instruments, as snubness is "not without" nose. So souls of plants and irrational animals cannot exist when separated from their bodies. The question whether any soul can so exist, and thus whether any soul is immortal (besides the souls of the heavens, which have immortal bodies), depends on whether all psychic powers are similarly dependent on bodily instruments. Sensation is not without its instruments, and imagination is not without sensation, so these are inseparable.

Some passages in De anima III suggest that intellection is not without imagination, so that it too is inseparable; other passages suggest that a special kind of intellection, of special matterless objects, is separable. (Fragments of Aristotle's "exoteric" works also argue that soul is immortal; perhaps Aristotle changed his mind from these early texts to the De anima, perhaps the texts can be reconciled, or perhaps the "exoteric" texts should be regarded as a popular approximation to a more precise truth.) De anima III, 5 says that "the passive nous is corruptible," and that only the active or productive nous is immortal. But what is this productive nous and what does it do? Since it is eternally and essentially intellectually cognizing, it seems that it must not be a part of the human soul, but rather a separate immaterial divine thing that acts on the "passive nous " in the soul. This recalls Platonic texts on nous (here best translated as "reason" or "rationality") as a separately existing virtue in which souls participate, the nous apparently personified as the divine craftsman of the Timaeus. Aristotle rejects all other separately existing virtues, because they are "not without" the irrational soul and the conditions of the body, but he has no reason to reject this one; and he too in Metaphysics Λ will identify such a nous with a world-governing divine archê.

For Aristotle, we can fully understand soul only by understanding its specific powers, their activities, and the objects and instruments of those activities; the De anima gives a general abstract account, which is filled in by the Parva naturalia, which treats of the actions and passions "common to soul and body"and almost all the soul's actions and passions are in common with the bodyand by the accounts of the instruments and activities of different animal species in the zoological works. But the neat sequence of "psychological works" (De anima and Parva naturalia ) followed by "zoological" or "biological" works (the History, Parts, Movement, Progression, and Generation of Animals ), as presented in Bekker and other modern editions, is probably an illusion. The texts themselves frequently refer to what has preceded or what will follow, but they seem to indicate two different sequences. Some texts, especially the Parts and Generation of Animals, imply a sequence in which the Parts would lead immediately into the Generation (both presupposing the History, as giving the facts for which they will supply the causes); the De anima and Parva naturalia would be a separate sequence, if anything more likely to come after than before (Aristotle refers to a lost part of the Parva naturalia, on the principles of health and disease, as the end point of natural philosophy).

But other texts imply a different order. Call "Parva naturalia Group I" the treatises connected with sensation, the On Sensation, On Memory, On Sleep, On Dreams, and On Divination in Sleep ; "Parva naturalia Group II" would be the On Length and Shortness of Life, On Youth, Old Age, Life, Death and Respiration, and the lost treatise on health. There are many indications for a sequence Parts of Animals, Progression of Animals, De anima, Parva naturalia Group I, Motion of Animals, Generation of Animals, Parva naturalia Group II, and perhaps a treatise on plants. It seems most likely that Aristotle began with the Parts-Generation sequence, and later inserted the other texts between the Parts and Generation, treating reproduction, like sensation and breathing, as an activity involving soul as well as body. No evidence supports putting the De anima before the Parts of Animals ; one option is to regard biology as beginning with the body, turning to the soul, and then exploring how they act together.


Sophia as an intellectual virtue"epistêmê and nous of what is most noble by nature"had been discussed in the Ethics. In the Metaphysics, Aristotle tries to provide a new discipline to bring us to this virtue, because he thinks that the existing disciplines with a claim to yield theoretical wisdomphysics, mathematics, and dialecticare insufficient. The awkward title, literally "The [books or things] after the physical [books or things]," first attested in Nicolaus of Damascus (1st century CE), reflects the difficulty of fitting the treatise into the standard scheme of disciplines: it belongs to theoretical philosophy, and draws on physics, but does not belong to physics, because the divine things it considers (unlike the heavens, also divine) exist separately from matter and motion.

The unity of the treatise is problematic. It is clear that Aristotle intended to write a long treatise on sophia, and that most of the books of the Metaphysics were intended as materials for such a treatise. But it is also clear (from almost verbatim duplication between Α9 and Μ45, verbatim duplication between the latter part of Κ and parts of the Physics, looser duplication between the former part of Κ and ΒΓΕ, and the coexistence of two books called alpha [now distinguished as Α and α]) that Aristotle never finished the treatise to his satisfaction. Perhaps he would have discarded some parts of the Metaphysics, and perhaps some were never intended for the treatise; and there are grounds for suspecting that Κ is a student's reworking of Aristotle's lectures. In what follows, it will be assumed that all the books except α and Κ were intended to belong to the treatise, but many scholars have doubted this for Δ and (less plausibly) Λ. (There is nothing to support the view, popular among nonspecialists, that the Metaphysics consists of fourteen independent books assembled by later Peripatetic editors; there are many forward and backward references between the books, including Δ and Λ.)

Aristotle says different things in different parts of the Metaphysics about the object of wisdom (or "first philosophy," as he says when distinguishing it from physics; once he calls it "theology"). Jaeger took these as evidence of different chronological strata: Aristotle would first have conceived the project of wisdom as searching for divine substances to replace the Platonic Forms, then reconceived it as a general study of being. But usually, and rightly, Aristotle's descriptions of wisdom are thought to be compatible. They are best taken as part of a developing strategy in the Metaphysics to narrow down and finally to acquire sophia.

It is perhaps most often thought that Aristotle aims at a universal science; that this project faces a difficulty, because "being" is said in different ways of things in different categories; that Aristotle proposes to solve this by discovering things that are in the primary way (these things, whatever they are, will be called substances, and once we understand their mode of being, we can understand the derivative modes of being of other things); that there are different and sometimes conflicting criteria for something to be in the primary way; that forms meet these different criteria better than matter or matter-form composites, but that the forms of corruptible things do so only imperfectly (because they are not separable except by reason); and that Aristotle therefore turns to divine forms (forms existing separately from matter), which will allow us to understand the derivative modes of being of other forms, other substances, and non-substances. This would explain why Aristotle can say that wisdom is about being, that it is about substance, and that it is about divine things.

However, the Metaphysics does not actually follow this program, and Aristotle nowhere calls divine things "forms," and nowhere says that they are beings or substances in any stronger sense than ordinary form-matter composites are (still less does he use them to understand the inferior modes of being of other things). Instead, Aristotle begins with an ethical characterization of wisdom, infers that wisdom will be a science of the archai (the "principles," or first of all things) and of first causes, then specifies these as causes of being, then reaches an account of divine things as archai and first causes of being, not as instances of a special sense of being. Theology is not a means to ontology; rather, ontology is a means to theology, or more precisely to "archeology" (knowledge of the archai might still count as wisdom even if there were nothing divine to know).

Metaphysics Α begins by characterizing wisdom as the kind of knowledge intrinsically most worth having, setting aside practical consequences; Aristotle then argues that this is knowledge of the archai, and that these archai will be first causes of all things. Indeed, all philosophers who believe in theoretical wisdom claim knowledge of some archai ; for pre-Socratic physicists, these are whatever existed from eternity before the ordered world arose out of them; for Platonic dialecticians, the Forms, especially maximally universal forms like being and unity; for Pythagorizing mathematicians, the one and the two or the infinite. We cannot directly observe any of these claimed archai, but must infer them as causes of more manifest things. Aristotle asks how each philosopher uses his archai as causesthat is, how the things he posits at the beginning of his account function in explaining the things he describes as arising later.

The best earlier philosophers, Anaxagoras and Empedocles and Plato, agree that among the archai is a Good and cause of goodness to the world. But, Aristotle claims, Anaxagoras and Empedocles cite only material and efficient causes (using nous or Love, their good archai, as efficient causes), and Plato cites only material and formal causes (using the one, his good archê, as a formal cause). Aristotle's main point is not that earlier philosophers have been discovering the four causes of his Physics, but that no one has yet used the Good as a final cause, thus no one has made it a cause whose goodness is explanatory. Aristotle thus motivates a new search for archai which will lead, in Λ, to a good archê as a final cause. He thus hopes to vindicate a key aspiration of Platonism, which Plato had undermined in his lecture on the Good by reinterpreting the Good as mathematical unity. Aristotle's rival Speusippus had concluded that the archê is One but not good; Aristotle makes the opposite decision, to discard the mathematics and save goodness. (Aristotle gives detailed objections against Academic accounts of Forms and numbers and their archai in Metaphysics ΜΝ.)

Metaphysics Β raises a series of aporiai, some about how the science of archai should proceed, some about the archai themselves, some about what things exist "by themselves" or as substances. If some X (a genus or a number, or being or unity) is not a substance, but is merely an attribute of some other underlying nature, then X is posterior to that nature and cannot be among the archai that wisdom seeks. Now while we know that the archai will be first causes, this does not tell us how to find them, since there are different kinds of causes, and different effects we might seek to explain.

Metaphysics Γ proposes to find the highest causes as causes of the most universal effects: "there is a science of being, inasmuch as it is being, and its per se attributes"a science that knows the causes to all things of the facts that they are, that they are each one, are severally many, and so on. It is sufficient to study the causes of being to substances, since the being of accidents is dependent on that of substances. (Γ argues that this science will also give explanatory understanding of the principles of noncontradiction and excluded middle.)

Metaphysics Δ distinguishes different senses of "archê," "cause," "one," "being," and other terms necessary for the investigation. Δ7 argues that "being" is said in several ways: being in different senses will have different kinds of cause, and confusion will result if we look for causes of being without drawing the necessary distinctions.

Metaphysics ΕΖΗΘ investigate causes of being in these different senses. Ε1 sets out the program of looking for the archai as causes of being, and specifically for archai which will be eternally unmoved and exist separately (not as attributes of something else); physics fails to reach unmoved archai, and mathematics fails to reach separately existing archai, and a new discipline of first philosophy or "theology" is needed. This might be dialectic, if Plato were right that the formal causes of things were eternal and separate, but he is not; Ε1 argues that physics, not first philosophy, understands the formal causes of natural things. Ε23 investigate the causes of "being per accidens," and Ε4 the causes of "being as truth," both concluding that no science (and certainly not wisdom) deals with these causes; the serious possibilities are "being as said of the categories," primarily of substance and derivatively of accidents, treated in ΖΗ, and "being as actuality and potentiality," treated in Θ. Metaphysics Ι ("Iota") deals with causes of per se attributes of being such as unity, difference, and contrariety, arguing that these do not lead to a separately existing one-itself or first pair of contraries, but only to a unit or a contrariety within each genus.

Metaphysics Ζ examines the causes of being as said of substances and accidents, but quickly restricts itself to the primary case, causes of substance. Aristotle speaks interchangeably of "the cause of substance to X " and "the substance of X." The conventional translation "substance" for ousia (the nominalization of the verb "to be") obscures the point that the ousia of X is whatever answers the question "what is X ?".

There are several ways we might answer this question, notably by giving the subject of X (i.e., a Y such that Y is X : "what is Socrates?" "this flesh and these bones"), or by giving the essence of X (i.e., a Y such that X is Y, or such that for X to be is for it to be Y : "what is man?" "wingless biped animal"), or by giving some part of the essence of X, such as a universal or genus under which X falls. The ousia of X taken the first way is its material cause; the ousia of X taken the second way is its formal cause. A philosopher might hope to reach archai, eternal and prior to sensible things, by starting with some sensible substance and asking "what is it?" repeatedly, in one of these ways, until some ultimate answer is reached: this might be, as a material cause, atoms and the void, or earth, water, air, and fire, or the "receptacle" of the Timaeus ; or, as a formal cause, Platonic Forms, especially the genera and being and unity.

Ζ devotes much ingenuity to showing that these projects do not succeed; what a sensible substance is is most properly its form, not a separate eternal form but one that does not exist prior to the form-matter composite. Plato might argue that, if the composite X came to be, there must already have been a form or essence of X for the process of coming-to-be to aim at; Aristotle agrees, but argues that this is not a separate eternal form, but a form existing in a generator of the same species (e.g., for an animal, the father) or in the soul of the artisan who produced X. Aristotle also argues that if the parts of the essence mentioned in the definition of X (like three lines in the definition of triangle, or like the four elements in Empedocles' definition of blood as "earth, water, air, and fire in equal proportions," or like animal and biped in the definition of man) were archai existing in actuality prior to X, X would not be one thing but many things (thus, as a reductio ad absurdum of Plato, there would not be one Form, Man, but two Forms, Animal and Biped). This argument might seem to make definition impossible, since the definientia are supposed to be prior to the definiendum ; but Aristotle argues that they can be definitionally prior without being capable of separate existence. There is no Animal that is just animal, prior to the differentiae of animals: an actual animal is always a biped animal or a quadruped animal or the like, and the genus "animal" is merely a potentiality for these differentiae. Likewise, actual matter is always hot or cold, wet or dry, and the common matter that underlies all sensible changes is only a potentiality, not something actually existing prior to all sensible things. Metaphysics ΖΗ thus give rules for definition, with implications for understanding the relations between genus and differentia, universal and particular, form and matter; but the archai of these definitions are not the eternal, separately existing first things sought by wisdom (whether pre-Socratic physics or Platonic dialectic or Aristotelian first philosophy); prior in definition but not in separate existence, they are objects of Aristotelian physics.

Metaphysics Θ examines causes of being as actuality and as potentiality. A power or potentiality (dunamis ), whether an active power to produce X or a passive power to undergo or become X, is a cause of X 's existing potentially (dunamei ). Most of the archai of the physicists would be potentialities or potential causes. Thus the "seeds" in Anaxagoras's precosmic mixture can become plants and animals and their functional parts; Anaxagoras's nous, or the demiurge of the Timaeus, prior to the cosmos, can act to produce order, but are not yet doing so. But such causes explain only the potential existence of the cosmos, and give no sufficient reason why the active archê should begin to act on the passive archê. That the effect exists actually (energeiâ[i] ) requires an activity (energeia ) or an actual cause ("housebuilder" is a potential efficient cause, "housebuilder housebuilding" an actual efficient cause). Aristotle tries, both to extract general concepts of dunamis and energeia, and to argue that energeia is prior to dunamis : seeds are not prior to mature living things (since a seed exists dependently on a previous mature member of the species), and the archai in the strict sense, the first of all things, are not dunameis or potential causes, but energeiai or actual causes. Thus against (say) Anaxagoras's conception of the archai as temporally and narratively prior to the cosmos, the archai must from all eternity have been acting to produce the cosmos, so the cosmos too must have existed from eternity.

Metaphysics Λ pulls together the threads of ΖΗΘ and draws conclusions for what causal chains lead up from changing sensible things to separate eternal archai. There is no single separately existing matter of all changeable things, nor a single form even for all things in the same species. While the form of a natural composite substance does not exist before the composite, its generator, a previous mature member of the same species, does exist before; but this chain of efficient causes goes back ad infinitum, without leading to a separate eternal archê. But Aristotle argues (drawing on Physics VIII) that the eternal continuance and approximate periodicity of sublunar generation require a further cause: not simply the sublunar generators, but something eternal and perfectly regularnamely, the rotations of the heavensthat sets the precise time lengths that sublunar cycles aim to approximate. Especially the daily and yearly motions of the sun, yielding the cycle of the seasons, serve to regulate cycles of generation.

Furthermore, these eternally unchanging motions require eternally unchanging substances as their efficient causes. Aristotle accepts Anaxagoras's and Plato's description of the mover of at least the first motion, the daily rotation of the whole heaven, as nous. But, using the premise that the archê must be pure energeia, he critically examines earlier philosophers' descriptions of nous 's causality, rejecting anything that would imply dunamis or changeability. Notably, nous must always move the heavens in the same way, and it must not move them in such a way as to be reciprocally affected by them. "Purifying" the Anaxagorean and Platonic accounts in this way, Aristotle concludes that nous moves the heaven only by causing the heaven to know and desire it: Nous is an efficient cause only by being a final cause. (When the heaven desires its mover, what does it desire to do, and how does this explain its motion? It should, like humans, order its actions toward contemplating God; and presumably its eternally unchanging motion is the best available imitation of God's eternally unchanging energeia.)

The premise that this nous is pure energeia also allows Aristotle (drawing on De anima III on "active nous ") to "purify" earlier accounts of how it thinks and what it thinks. It is not a cognitive ability that could be applied to many objects, but a single eternal act of cognition of a single eternal objectthe best object, or "good-itself." If this object were outside nous, nous would depend on something external to complete its act of cognition, and would of its own essence be merely potential nous ; Aristotle concludes that nous is identical with the good-itself that it contemplates. This result allows Aristotle to fulfill various promises about wisdom from Metaphysics Α, showing how the good is a cause, qua good (as a final cause), and not just as an efficient or formal cause. He vindicates Plato's promise of a single first good archê against Speusippus's criticism, but only by giving up on talk of the One, and finding a causal route up to the archê from physics rather than from mathematics.


Aristotle's immediate influence came through the Peripatetic school, led after Aristotle's death by his student Theophrastus; other important students were Eudemus, Aristoxenus, and Dicaearchus. The Peripatetic Demetrius of Phalerum governed Athens, backed by Macedonian power, from 317 to 307 BCE. But the Peripatetic school declined after this (perhaps in part because of the reaction against Demetrius at Athens); for most of the Hellenistic period (32330 BCE), the dominant schools were the Academics, Stoics, and Epicureans. Peripatetics turn up more at Alexandria than at Athens, and more in biography and literary scholarship than in scientific philosophy.

However, there was a revival of Aristotle, as well as of Plato, in the first centuries BCECE, and attention turned from the "exoteric" texts to the "acroamatic" texts as offering a systematic teaching in all philosophical disciplines. Teaching would take place, by oral exposition of the texts of Aristotle, in whatever was thought to be the correct sequence, accompanied by refutations of more recent schools and solutions to new aporiai. This oral teaching is reflected in written commentaries, of which the most important are those of Alexander of Aphrodisias (c. 200 CE); we also have paraphrases of several Aristotelian treatises by Themistius (fourth century CE).

Besides the Peripatetics, late ancient Platonists make use of Aristotelian concepts in trying to extract a systematically teachable technical philosophy out of Plato's dialogues, and often wind up incorporating Aristotelian doctrines. In particular, they share Aristotle's concern to avoid inappropriately assimilating soul or nous or other divine realities to lower things, notably by attributing to them extension or change or dunamis. (Aristotle is here seen as an ally especially against Stoic corporealism.) Where Plato describes intelligible forms as conspecific with sensible things, the demiurge as acting after a period of inactivity, or thinking as a circular motion of soul, the Platonists use Aristotle's arguments, together with a principle of charity, to argue that Plato must have intended these comparisons to sensible things to be understood allegorically; they say either that Aristotle's criticisms of Plato are misunderstandings, or that Aristotle intended to criticize not Plato, but only disciples who took Plato's metaphors literally. At the same time, they reinterpret Platonic forms as a plurality of sciences in God, weakening Aristotle's insistence on the singleness of God's knowledge. Thus fifth- and sixth-century commentaries both on Plato and on Aristotle harmonize the two authors to some extent. This is taken furthest by Simplicius; by contrast, John Philoponus, for Christian reasons, defends some specifically Platonic doctrines, including creation in time, against Aristotle.

After the mid-sixth century, the teaching of philosophy collapses, beyond introductions to philosophy and lectures on Porphyry's Introduction (Isagoge ) to the Organon and Aristotle's Categories, De interpretatione, and the elementary part of the Prior Analytics. The recovery of the rest of Aristotle's work as a basis for systematic philosophical instruction occurred first in the Islamic world; key figures are al-Fārābī and Avicenna (Ibn Sīnā). These thinkers accept the guidance of late ancient commentators, and thus share to some extent in the harmonizing of Aristotle and Plato; by contrast, Averroes (Ibn Rushd) champions Alexander of Aphrodisias against harmonistic commentators, and tries to defend "scientific" Aristotelian philosophy against what he sees as unscientific Platonist contamination. Versions of Avicennian philosophy are taught in Iran to the present day.

In Greece, Michael Psellus revived late ancient Platonic-Aristotelian philosophy, which remained vital until the fall of the Byzantine Empire. In the Latin West, knowledge of Aristotelian philosophy survived in the translations and commentaries on the Organon by Boethius; Abelard and his twelfth century contemporaries began a renaissance of Aristotelian philosophy based almost wholly on the logic. Around 1200, translations of the whole Aristotelian corpus became available (first from Arabic, along with Arabic commentaries and treatises, then directly from Greek), and a systematic teaching of Aristotelian philosophy ("scholasticism") became the basis for university instruction, and a prerequisite for the study of Christian theology; different solutions were proposed to the conflicts between Aristotle and biblical revelation (key figures are Thomas Aquinas, John Duns Scotus, and William of Ockham).

While scholastic Aristotelianism flourished in the Renaissance (key figures are Pietro Pomponazzi in Italy, Francisco Suárez in Spain and Portugal), there is also much Renaissance polemic against Aristotle. The charge may be that he is irreligious (he makes the causal connection of God with the world too thin, and seems to deny providence over the sublunar world and the immortality of human souls; he certainly denies miracles such as creation in time or resurrection); that his claims of scientific knowledge cannot overcome skeptical challenges; or that his explanations are tautologous, multiplying words without practical consequences either technical or moral. These criticisms are taken up by the mechanical philosophers of the seventeenth century (notably Descartes, Gassendi, Hobbes), who aim to give a systematic replacement for Aristotelian physics, doing without forms or qualities superadded to matter (except possibly the human rational soul), abolishing the distinction between heavenly and earthly matter, and deriving phenomena from a natural tendency of bodies to persist in rectilinear motion, and from the results of collisions between bodies. Since the successes of this new physics (culminating in Newton), no systematic revival of Aristotelian philosophy has been possible; likewise, modern mathematical logic has permanently eclipsed Aristotelian syllogistic. Kantians often accuse Aristotle of uncritical realism in epistemology.

But Aristotle continues to be central in philosophical education, and to be a source of inspiration, chiefly in practical philosophy, metaphysics, and the philosophy of mind. Often philosophers have turned back to Aristotle for a description of phenomena of ordinary experience and language, careful attention to which (it is claimed) would undermine the appeal of oversimple modern reductionist theories (utilitarianism, associationism, materialism), without positing anything radically beyond ordinary experience (categorical imperatives of pure reason, intellectual intuitions, incorporeal substances). Neo-Aristotelians prefer intensional to extensional distinctions: a soul is not a substance other than the living body, but is the body itself qua living and not merely qua body. And these intensional differences are discerned, not by intellectual intuition, or by Kantian a priori synthesis, but by ordinary perception disciplined to recognize things as what they are. (Thus the practical rationality required for virtue is neither means-end reasoning nor a Kantian faculty of rules, but a sensitivity to morally salient features of situations.) Aristotle is seen as seeking a "middle way," for example, between pre-Socratic materialism and Platonic metaphysics, that could be a model for modern philosophers. Such interpretations tend to understate the commitments that Aristotle shares with Plato, and his internal criticisms and refinements of the Platonic philosophical (and theological) project; the use of Aristotle for inspiration in contemporary philosophy should be balanced by an awareness of the risks of removing Aristotle from his context and reducing him to what seems usable for current philosophical problems.

See also Anaxagoras of Clazomenae; Ancient Aesthetics; Aristotelianism; Logic, History of: Ancient Logic; Neoplatonism; Plato; Theophrastus.


There is an excellent annotated bibliography in Jonathan Barnes, ed., The Cambridge Companion to Aristotle (Cambridge, U.K.: Cambridge University Press, 1995); there are also comprehensive bibliographies in Hellmut Flashar, ed., Grundriss der Geschichte der Philosophie ("Ueberweg"), Die Philosophie der Antike, Vol. 3, pp. 407492 (Basel: Schwabe, 2004), and, for the Metaphysics, in Robero Radice and Richard Davies, Aristotle's Metaphysics: Annotated Bibliography of the Twentieth-Century Literature (Leiden: Brill, 1997). Only primary sources, indispensable scholarly tools, and starting points for entering into scholarly discussions, including some particularly important publications that have appeared since Barnes, will be listed here. All the transmitted works and many fragments are translated in Jonathan Barnes, ed., The Complete Works of Aristotle, 2 vols. (Princeton, NJ: Princeton University Press, 1984); the translations are not uniform, and some are better than others. Many works have been well translated, with useful notes, in the Clarendon Aristotle series from Oxford University Press (thus far: Categories and De interpretatione, translated by J. L. Ackrill, 1962; Posterior Analytics, translated by Jonathan Barnes, 2nd ed., 1994; Topics I and VIII, translated by Robin Smith, 1994; Physics III, translated by William Charlton, 1980; Physics IIIIV, translated by Edward Hussey, 1983; Physics VIII, translated by Daniel Graham, 1999; De generatione et corruptione, translated by C. J. F. Williams, 1982; De anima IIIII, translated by D. W. Hamlyn, 2nd ed., 1993; De partibus animalium I and De generatione animalium I, translated by David Balme, 2nd ed., 1993; De partibus animalium, translated by James Lennox, 2001; Metaphysics Β, translated by Arthur Madigan, 1999; Metaphysics ΓΔΕ, translated by Christopher Kirwan, 2nd ed., 1993; Metaphysics ΖΗ, translated by David Bostock, 1994; Metaphysics ΜΝ, translated by Julia Annas, 1977; Eudemian Ethics I, II, and VIII, translated by Michael Woods, 2nd ed., 1992; Nicomachean Ethics VIIIIX, translated by Michael Pakaluk, 1998; Politics III, translated by Trevor Saunders, 1996; Politics IIIIV, translated by Richard Robinson, 1962; Politics VVI, translated by David Keyt, 1999; Politics VIIVIII, translated by Richard Kraut, 1997). The editions and translations in the Loeb Classical Library are usually less good. Hackett has published good annotated translations of the Prior Analytics (translated by Robin Smith, 1989), Nicomachean Ethics (translated by Terence Irwin, 1985), Politics (translated by C. D. C. Reeve, 1998), and Poetics (translated by Richard Janko, 1987), all Indianapolis, IN: Hackett. Also useful is George Kennedy's annotated translation of the Rhetoric, On Rhetoric: A Theory of Civic Discourse, New York: Oxford University Press, 1991. Many works are also available in the Budé series, from Editions Les Belles Lettres, Paris: Presses Universitaires de France, Greek text with facing French translation, and in the Oxford Classical Texts (Greek only, many of them edited by W. D. Ross) and the Teubner series (Greek text only; previously Leipzig: Teubner or Stuttgart: Teubner, now Munich: Saur). W. D. Ross's editiones maiores (Greek text with English introduction and commentary) of the Metaphysics (2 vols., 1924), Physics (1946), Analytics (1955), De anima (1961) and Parva naturalia (1955), and Harold Joachim's editio maior of the On Generation and Corruption (1922), all Oxford: Oxford University Press, are extremely useful. However, on the Metaphysics and De anima, it is advisable to compare Werner Jaeger's Oxford Classical Text (1957) and Antonio Jannone and Edmond Barbotin's Budé (1966), respectively; with Joachim compare Marwan Rashed's Budé (2005). The Budé texts are generally more prudent and reliable than Ross's; some have very useful introduction and notes, others are only sparsely annotated. For the Categories, compare the Oxford Classical Text of Lorenzo Minio-Paluello (also including the De interpretatione, 1949) with Richard Bodéüs' Budé (2001), which puts "[Catégories ]" in brackets on the title page and then switches to Les Avant les Lieux ("The Before the Topics"). Jacques Brunschwig's Budé Topiques (Vol. 1, 1967; Vol. 2, forthcoming) particularly stands out. For the Rhetoric, the standard text is Rudolfus Kassel, Aristotelis Ars rhetorica, Berlin: De Gruyter, 1976. For the De motu animalium, see Martha Nussbaum, Aristotle's De motu animalium, Princeton, NJ: Princeton University Press, 1978; for the Historia animalium, see David Balme, Aristotle: Historia animalium, Cambridge, U.K.: Cambridge University Press (Vol. 1, 2002; Vol. 2, forthcoming). For the fragmentary Protrepticus, compare Ingemar Düring, Aristotle's Protrepticus: An Attempt at Reconstruction, Göteborg: Acta Universitatis Gothoburgensis, 1961 (including an edition of the fragments with English translation) with the forthcoming translation by Douglas Hutchinson and Monte Johnson (expected Indianapolis: Hackett). For the fragmentary On Ideas, see Gail Fine, On Ideas, Oxford: Oxford University Press, 1993, and Walter Leszl, Il "De ideis" di Aristotele e la teoria platonica delle idee, Florence: Olschki, 1975 (including a critical edition by Dieter Harlfinger). An extremely useful tool for studying concepts across the Aristotelian corpus (unfortunately presupposing Latin as well as Greek, and using a reference system that takes some getting used to) is Hermann Bonitz, Index Aristotelicus, Berlin: Reimer, 1870.

The most important secondary literature on Aristotle remains the late ancient commentaries collected in Commentaria in Aristotelem Graeca, 23 vols., Berlin: Reimer, 18821909, especially those of Alexander of Aphrodisias and Simplicius. Many of these volumes have been translated into English in the series Ancient Commentators on Aristotle, edited by Richard Sorabji, London: Duckworth, and Ithaca, NY: Cornell University Press. On the Greek commentators see Richard Sorabji, ed., Aristotle Transformed, London: Duckworth, 1990, and Richard Sorabji, ed., Philoponus and the Rejection of Aristotelian Science, London: Duckworth, 1987. For the earlier Greek reception of Aristotle, see Paul Moraux, Der Aristotelismus bei den Griechen, 2 vols., Berlin: De Gruyter, 19731984. For the availability of medieval Arabic and Latin commentaries on Aristotle, see the articles on, especially, al-Fārābī, Averroes (Ibn Rushd), and Thomas Aquinas.

Articles on Aristotle, edited by Jonathan Barnes, Malcolm Schofield, and Richard Sorabji, 4 vols., London: Duckworth, 19751979, is a good cross section of modern scholarship up to its time, with non-English articles translated. There are also excellent collections of articles, usually on a particular work of Aristotle, often arranged so as to form a collective commentary, in the proceedings of the triennial Symposia Aristotelica, including thus far: Ingemar Düring and G. E. L. Owen, eds., Aristotle and Plato in Mid-Fourth Century, Göteborg: Acta Universitatis Gothoburgensis, 1960; Suzanne Mansion, ed., Aristote et les problèmes de méthode, Louvain: Publications universitaires, 1961; G. E. L. Owen, ed., Aristotle on Dialectic: The Topics, Oxford: Oxford University Press, 1968; Ingemar Düring, ed., Naturphilosophie bei Aristoteles und Theophrast, Heidelberg: Stiehm, 1969; Paul Moraux, ed., Untersuchungen zur Eudemischen Ethik, Berlin: De Gruyter, 1971; Pierre Aubenque, ed., Etudes sur la Métaphysique d'Aristote, Paris: Vrin, 1979; G. E. R. Lloyd and G. E. L. Owen, eds., Aristotle on Mind and the Senses, Cambridge, U.K.: Cambridge University Press, 1978; Enrico Berti, ed., Aristotle on Science: the Posterior Analytics, Padova: Editrice Antenore, 1981; Paul Moraux and Jürgen Wiesner, eds., Zweifelhaftes im Corpus Aristotelicum, Berlin: De Gruyter, 1983; Andreas Graeser, ed., Mathematics and Metaphysics in Aristotle (= Metaphysics ΜΝ), Bern: Haupt, 1987; Günther Patzig, ed., Aristoteles' Politik, Göttingen, 1990; David Furley and Alexander Nehamas, eds., Philosophical Aspects of Aristotle's Rhetoric, Princeton, NJ: Princeton University Press, 1994; Michael Frede and David Charles, eds., Aristotle's Metaphysics Lambda, Oxford: Oxford University Press, 2000; and Frans de Haas and Jaap Mansfeld, eds., Aristotle's On Generation and Corruption I, Oxford: Oxford University Press, 2004.

The classic attempt to write Aristotle's intellectual biography is Werner Jaeger, Aristotle: Fundamentals of the History of his Development, translated by Richard Robinson, Oxford: Oxford University Press, second edition, 1948. For a survey of the status quaestionis on Aristotle's development, see William Wians, ed., Aristotle's Philosophical Development, Lanham, MD: Rowman and Littlefield, 1996. The ancient sources on Aristotle's life are collected by Ingemar Düring, Aristotle in the Ancient Biographical Tradition, Göteborg: Acta Universitatis Gothoburgensis, 1957. On the titles and histories of Aristotle's writings, see Paul Moraux, Les listes anciennes des ouvrages d'Aristote, Louvain: Editions universitaires de Louvain, 1951; on stories about the rediscovery of the texts and on the possible role of Andronicus of Rhodes in editing and ordering Aristotle's texts, compare the treatment of Paul Moraux, Der Aristotelismus bei den Griechen (cited above) with that of Jonathan Barnes, "Roman Aristotle," in Philosophia Togata II, edited by Jonathan Barnes and Miriam Griffin, pp. 169, Oxford: Oxford University Press, 1997; and see Stephen Menn, "The Editors of the Metaphysics," in Phronesis 40, no. 2 (July 1995): 202208. On Aristotle's treatment of his predecessors, the classic (but highly controversial) works are Harold Cherniss, Aristotle's Criticism of Presocratic Philosophy, Baltimore, MD: Johns Hopkins Press, 1935, and Aristotle's Criticism of Plato and the Academy, Vol. 1 (Vol. 2 was never published), Baltimore, MD: Johns Hopkins Press, 1944.

On the Nicomachean Ethics, starting points are R. A. Gauthier and J. Y. Jolif, L'Ethique à Nicomaque, introduction, traduction et commentaire, 2 vols. in 3, Louvain: Publications universitaires de Louvain, 19581959; W. F. R. Hardie, Aristotle's Ethical Theory, 2nd ed., Oxford: Oxford University Press, 1980; Sarah Broadie, Ethics with Aristotle, Oxford: Oxford University Press, 1991; and Amélie Rorty, ed., Essays on Aristotle's Ethics, Berkeley: University of California Press, 1980. On the relation between the different ethical works, starting points are (besides Michael Woods's Clarendon Eudemian Ethics I, II, and VIII) Anthony Kenny, The Aristotelian Ethics, Oxford: Oxford University Press, 1978 (although Kenny's claim that the Nicomachean Ethics is earlier than the Eudemian is idiosyncratic); and John Cooper, "The Magna Moralia and Aristotle's Moral Philosophy," in his Reason and Emotion, pp. 195211, Princeton, NJ: Princeton University Press, 1999. On the Politics, besides the introductory volume and the essays and notes of the classic W. L. Newman, The Politics of Aristotle, 4 vols., Oxford: Oxford University Press, 18871902, see now Richard Kraut, Aristotle: Political Philosophy, Oxford: Oxford University Press, 2002; and, for the context of political debate, Josiah Ober, Political Dissent in Democratic Athens, Princeton, NJ: Princeton University Press, 1998. On the Constitution of the Athenians, see P. J. Rhodes, A Commentary on Aristotle's Athênaiôn Politeia, Oxford: Oxford University Press, 1981; and John Keaney, The Composition of Aristotle's Athênaiôn Politeia, New York: Oxford University Press, 1992.

On dialectic, besides Robin Smith's Clarendon Topics I and VIII, the introduction to Jacques Brunschwig's Budé Topiques, and G. E. L. Owen's Symposium Aristotelicum volume, all cited above, see the introduction to Eleonore Stump, Boethius De differentiis topicis, Ithaca, NY: Cornell University Press, 1978; and Robin Smith, "Aristotle on the Uses of Dialectic," in Synthese 96 (1993): 335358. The majority of English-language literature on dialectic is really about the use of pre-reflective intuitions or ordinary language in argument, and has little to do with what Aristotle calls dialectic. On the Categories, see Michael Frede, "Title, Unity and Authenticity of the Aristotelian Categories," in his Essays on Ancient Philosophy, pp. 1128, Oxford: Oxford University Press, 1987; and Stephen Menn, "Metaphysics, Dialectic, and the Categories," in Revue de Métaphysique et de Morale 100, no. 3 (JulySeptember 1995): 311337; on the On Interpretation, see C. W. A. Whittaker, Aristotle's De interpretatione: Contradiction and Dialectic, Oxford: Oxford University Press, 1996; on the Sophistical Refutations, see Louis-André Dorion, Aristote, Les refutations sophistiques, introduction, traduction et commentaire, Paris: Vrin, 1995. On Aristotle's syllogistic, besides Robin Smith's introduction to his Hackett Prior Analytics, see Jan Łukasiewicz, Aristotle's Syllogistic, 2nd ed., Oxford: Oxford University Press, 1957; Günther Patzig, Aristotle's Theory of the Syllogism, Dordrecht: Reidel, 1968; and Paul Thom, The Syllogism, Munich: Philosophia, 1981; on demonstration, see Enrico Berti's Symposium Aristotelicum volume (Aristotle on Science ) and Jonathan Barnes' Clarendon Posterior Analytics. On the Rhetoric, the Symposium Aristotelicum volume (D. Furley and A. Nehamas, Aristotle's Rhetoric ) gives one approach, and George Kennedy's translation (mentioned above), and his The Art of Persuasion in Greece, Princeton, NJ: Princeton University Press, 1963, give another. On the Poetics, good starting points are Richard Janko's Hackett and Amélie Rorty, ed., Essays on Aristotle's Poetics, Princeton, NJ: Princeton University Press, 1982.

On physics and cosmology, there is unfortunately not much to recommend, beyond the editiones maiores and Symposium Aristotelicum volumes noted above; still useful is Friedrich Solmsen, Aristotle's System of the Physical World, Ithaca, NY: Cornell University Press, 1960; on Aristotle on causality see Richard Sorabji, Necessity, Cause and Blame, London: Duckworth, 1980; and for the issues of Greek cosmology, see David Furley, The Greek Cosmologists, Vol. 1, Cambridge, U.K.: Cambridge University Press, 1987 (the promised Vol. 2 has not yet appeared), and his collection of articles Cosmic Problems, Cambridge, U.K.: Cambridge University Press, 1989. By contrast, there is a large literature on Aristotle's psychology and zoology. For orientation to the literature on the psychology, see Martha Nussbaum and Amélie Rorty, eds., Essays on Aristotle's De anima, 2nd ed., Oxford: Oxford University Press, 1995; and see Stephen Menn, "Aristotle's Definition of Soul and the Programme of the De anima," in Oxford Studies in Ancient Philosophy, 22 (Summer 2002): 83139. On Aristotle's ordering and reordering of the psychological-physiological-zoological corpus, see Marwan Rashed, "Agrégat de parties ou vinculum substantiale ? Sur une hésitation conceptuelle et textuelle du corpus aristotélicien," in André Laks and Marwan Rashed, eds., Aristote et le mouvement des animaux: dix études sur le De motu animalium, pp. 185202, Lille: Presses universitaires du Septentrion, 2004. On zoology, besides the translations and commentaries of Lennox, Balme, and Nussbaum, cited above, the best starting point is Alan Gotthelf and James Lennox, eds., Philosophical Perspectives on Aristotle's Biology, Cambridge, U.K.: Cambridge University Press, 1985.

For the Metaphysics, it is always helpful to compare Theophrastus's Metaphysics, for which see the Budé by André Laks and Glenn Most, 1993; for an English translation, see W. D. Ross and F. H. Fobes, eds., Theophrastus, Metaphysics, Oxford: Oxford University Press, 1929. Modern landmark studies on the Metaphysics are Werner Jaeger, Studien zur Entstehungsgeschichte der aristotelischen Metaphysik, Berlin: Weidmann, 1912; Pierre Aubenque, Le problème de l'être chez Aristote, Paris: Presses universitaires de France, 1966; Joseph Owens, The Doctrine of Being in the Aristotelian Metaphysics, Toronto: Pontifical Institute of Medieval Studies, 3rd ed., 1978; Günther Patzig, "Theology and Ontology in Aristotle's Metaphysics " in Articles on Aristotle (cited above), Vol. 3, pp. 3349; Vianney Décarie, L'objet de la métaphysique selon Aristote, Paris: Vrin, 1961; Giovanni Reale, The Concept of First Philosophy and the Unity of the Metaphysics of Aristotle, translated by John Catan, Albany: SUNY Press, 1980; Michael Frede, "The Unity of General and Special Metaphysics: Aristotle's Conception of Metaphysics," in his Essays on Ancient Philosophy (cited above), pp. 8195; Michael Frede and Günther Patzig, Aristoteles: Metaphysik Ζ: Text, Übersetzung, Kommentar, Munich: Beck, 1988; and Myles Burnyeat, A Map of Metaphysics Zeta, Pittsburgh: Mathesis Publications, 2001.

Stephen Menn (2005)