1.    On science and scientists

The essential difference between science and related forms of intellectual activity is very difficult to define exactly. A dictionary definition of science is an accurate reflection of what some people understand by the term ‘science’, and then only in the period and place in which the dictionary was written. Most people live their lives and use their language without reference to a dictionary. Even sticking to the dictionary definition, what the word ‘science’ has meant to different people has varied over time and cultures, and its meaning continues to change¹. As such, the term itself can be a source of anachronism in the study of science in ancient times. For we naturally tend to find in ancient authors just those sorts of thing which we recognize as science in our own times, and we tend to ignore those things which are incomprehensible or just plain wrong–to our way of thinking. And we naturally tend to organize what we find into categories which reflect our way of dividing up the world into subjects and disciplines. Thus things that the ancients linked together, we tear apart and treat separately. In particular, some we include in the category of science, others we exclude.

In this process, Crombie has emphasized that the circumscription of subject matter –what is regarded as science, what counts as valid questions, pieces of evidence, types of argument and answer – is the key to opening doors to new knowledge and understanding; but that at the same time it closes other doors. For example, modern astronomers are expected not to engage in astrology, but that door remained open for their predecessors, from ancient Greece to enlightenment Britain, who were socially and intellectually free to study both subjects simultaneously. The ancients put astrology on a par with astronomy, as two sides of the same coin². Changing views of the meaning of ‘science’ closed this door in due course, and classical scholarship felt its influence. So, until very recently, modern scholars have highlighted the ancients’ achievements in ancient astronomy, and done their best to ignore (if not positively conceal, excuse, or try to explain away) what they regard as an embarrassing lapse of rationality³.

Modern intellectual traffic between the ‘science’ end of ancient activities and the ‘superstition’ end of the same, even when both were conducted by the same ancient person, tends to be one-way. Astronomy-historians, for example, can produce works focused tightly on ancient theories of the cosmos. They may extend this to include closely related matters, such as the business of actually observing heavenly phenomena, but it is not considered necessary for an intellectually sound treatment of ancient astronomy. Astrology-historians need to explain ancient astronomical theories before they can start on the real subject of their interest, which is astrological theories. And alchemy-historians need to know ancient astrology (and therefore ancient astronomy) in order to deal with their subject of interest⁴. But a historian of astronomy considers alchemy well outside the remit of his study.

For the purpose of these pages, I too have divided the subject matter into modern categories, in spite of what I have said above, for three pragmatic reasons. First, because the original ancient categories are too large to serve as analytical tools. For the Greeks, all serious intellectual pursuits could be classified under the head philosophia. The Greek term usually translated as ‘science’, episteme, covered cooking, amongst other things⁵; with due respect to the modern category of domestic science, that will not do for our purposes. Natural history, or more precisely, inquiry into phusis, nature, was the nearest phrase to the modern term ‘science’, but if we use this category, then we cannot sort the material, because almost everything falls into the one category, and the only obvious tool left to apply to break the mass up into manageable chunks is the chronological one, loosely applied⁶. As W Knorr pointed out in a mathematical context⁷, ‘to explain certain aspects of ancient geometry, it may become advisable, even necessary, to import notions from more recent fields, like algebra. This raises the possibility of anachronism, as is present in all analogical forms of exegesis. That risk becomes acceptable if the alternative is the learner’s incomprehension.’ I have taken that risk by using modern categories. A second reason why I have divided the subject matter into modern categories is because most of the existing modern literature focuses on one or another category, and since this a survey, it seems sensible to organize it on the same principles. The third reason is because a brief survey of a vast field is not the place to try to develop a new taxonomy of ancient science. However, within each category it is my intention to include material which sits uncomfortably within it, simply in order to make clear the limitations of these categories.

A separate and serious problem with ‘science’ as a category is that the abstract nature of scientific ideas seems to encourage an abstraction in the study of its history. Often the ideas are presented in splendid isolation from the social, political, religious and economic context of their formation⁸. (This problem is not confined to ancient history.) For example, when one reads the word ‘school’ applied to e.g. Aristotle’s Lyceum, one imagines a building, at the very least. Yet there was no private building until after Aristotle’s death⁹; his ‘colleagues’ and ‘students’ met him and each other in a sacred public wood outside Athens’ city walls, where ho boulomenos¹⁰ could worship Apollo Lyceus, or go for a stroll, or collect firewood, or hunt hares, or whatever¹¹. Meanwhile ancient studies carried out by people now more or less anonymous in circumstances not fitting with modern notions of research can be overlooked. For example, Mithridates’ immunity to all known venoms and poisons¹² was the result of a long programme of research and development by toxicologists who worked in his service¹³, combined with empirical trials on death-row prisoners in the pursuit of antidotes. Pompey (at least) was interested in this scientific project: he ordered a Latin translation of Mithridates’ written results (Pliny NH 25.7). These results, like those obtained by the human vivisectionists of the Hellenistic period, were incorporated into the ancient scientific knowledge base, and can now be studied in comfortable abstraction from the context of their creation. Medical ethics is not new incidentally, and, as with vegetarianism, the arguments have not changed much since antiquity.

One of the major features of Greek science is that most of its practitioners were autodidacts. Even those who studied under a philosophical giant seem, with very few exceptions, not to have been content to follow a path laid down by a predecessor. They wished to carve out their own path, citing predecessors’ views when it suited them, and ignoring them likewise. They did not so much stand on their predecessors’ shoulders as knock them down, step over them, and go elsewhere¹⁴. In this respect, the nature of the scientific endeavour today is quite different from antiquity. A big and related difference, with many profound corollaries, is the individualism of ancient science. Though we (and the ancients) occasionally talk of philosophical ‘schools’ or medical ‘sects’, these terms more often refer to like-minded individuals, who might be widely dispersed over time and space, than to organizations of intellectual co-workers and colleagues. Someone who never set foot in the Akademy might be called a Platonist, because the label reflected his philosophical inclinations and not his membership of some club, society or association.

Moreover, it is a large and usually overlooked assumption that the authors of ancient scientific texts were ‘professional’ philosophers or doctors or teachers or people otherwise engaged most of their time in thinking or writing about their subject. The same point can be made for other producers of literary works, e.g. dramatists¹⁵. It seems to me probable that a significant number of these people undertook their studies in their (relatively ample) leisure time. As late as the last century, the men who founded ancient history and kindred subjects founded them in their spare time: for example, George Grote was a banker, MP and a few other things simultaneously, Thomas Heath was a civil servant and James Gow was a lawyer; they were not salaried academics.

For much of Western history there is an assumption that writers belong to an educated elite. But this assumption is highly presumptuous for ancient Greece. A fair number of the Greek authors were, we can safely assume, born to farm¹⁶, whilst others were taught craft skills by their fathers. Sokrates’ dad was a stonemason; Aristotle’s a healer; Theophrastos’ a laundryman; Ktesibios’ a barber. Apollonius of Perga was called a carpenter¹⁷; several creators of panaceas mentioned by Galen are labelled tradesmen (or worse) of one sort or another¹⁸; and Archimedes was apparently at Hieron's beck and call¹⁹. Demetrius of Phaleron governed Athens before organizing the Library at Alexandria²⁰; a number of Greek intellectuals appeared on the Roman scene first as civic officials, ambassadors for their poleis²¹; Frontinus was consul and governor of Britain (amongst other things), whose interest in hydraulic engineering arose from his (important political) appointment by Nerva to the office of overseer of the water supply to the city of Rome²²; and Pliny was Commander of the Fleet at Misenum when he died, famously observing Vesuvius from too close quarters. If this point is not convincing, consider, as the implied alternative, the plausibility of an ancient Greek boy (still less a Roman) telling his father that he wanted to be a professional philosopher when he grew up²³.

Contextualisation will not just benefit historians of science. ‘To study what passes for science in a society is to go to the centre of the values of that society’, as Lloyd has pointed out forcefully²⁶. The ancient historian with little or no interest in ancient science can learn a great deal about ‘ordinary’ history from the scientific texts produced in any particular place and time. They contain a large amount of incidental information concerning everyday life and ideas at the time the texts were written. Nutton (1972 p. 50) described Galen’s Prognosis as ‘the finest contemporary account of society in Rome during the reign of Marcus Aurelius’. Consider a selection of ‘scientific’ texts from just one century (the fourth BC), and from one place (Athens): in Theophrastos’ treatise On Fire, one discovers that there were entertainers eating fire and performing the fire-walk; from his Metaphysics that puppies were kept in quail cages to produce small dogs; from his On Stones that coal was mined and used as fuel; and from the Aristotelian Problems that sponge divers were using diving bells. None of this is obvious from the dramatists, historians, orators, novelists, large pile of inscribed stones, and other sources typically used by ancient historians. The primary sources for any period consist not just of the well known and well ploughed texts: there are a lot of grossly underutilised ‘scientific’ texts too, which cast a different and sometimes brilliant light on ancient society.

This incidental information is also important to the historian of science, for the society in which the science was created shapes the science itself, and neither can be properly understood in isolation from the other. This will be a theme throughout these pages.

2.    On the methods and assumptions of Greek science

Geoffrey Lloyd, more than any other modern scholar, has highlighted the variety of theories, ideas, and opinions which co-existed in ancient Greece, and their interdependence. We moderns may, quite naturally, alight on one or two which seem to be ‘on the right track’ or to be ‘predecessors’ of modern theories or methods, but the ancients did not know which way natural history or medicine or philosophy or technology would develop after them; they did not know which theories were ‘right’ – or at any rate ‘less wrong’ (as judged by today’s beliefs and standards); and they did not know which theories would ‘fit’ into a modern concise history of a particular discipline. They did not, in short and of course, have hindsight.

As a result, ancient society as a whole did not give the kind of precedence we might give to one theory or sect over another: Hippocratics over herbalists, for example. If we write histories which focus only on those theories which complement or are consistent with modern theories (as was usual for previous generations of scholars), and ignore those ancient theories which can be so inconsistent with modern thinking that it is almost embarrassing to discuss them (akin to drawing attention to a friend’s faux pas), then our histories will give a very distorted view of ancient science and ancient society.

Ancient theories developed in competition with one another. The agonistic nature of Greek culture penetrates their science too. Developing as a response to a pre-existing theory, modified in response to new theories or new arguments or new evidence, shaped by those debates rather than by a systematic and ‘objective’ programme of research into some type of natural phenomenon, the various theories should not be seen as free-standing intellectual systems developed in an intellectual vacuum. Their content and form is shaped partly by the content and form of other, pre-existing and contemporaneous, theories and ideas. To quote Lloyd (1989, pp. 431-2), ‘[the Greeks] created, they invented, their own distinctive and divergent ideas, often in direct and explicit confrontation with their rivals. The concept of nature was forged in controversy, notably as the underpinning to the claims made by new styles of wisdom in their attempts to outbid more traditional kinds.’ No area of ancient scientific work was free from such dispute, even Euclidean geometry²⁷.

Lloyd is also responsible for advancing the idea that the Greek ‘scientific’ spirit grew out of the socio-political environment, where decisions were made by debate not by fiat, in non-monarchical systems of government, where an idea was tested by reasoning and not accepted as given. This idea²⁸ has been picked up widely by historians of Greek science, and has much to commend it as an explanatory factor for the extraordinary emergence of ‘scientific thinking’. It does not dove-tail neatly with the standard view on socio-political development in the Greek poleis²⁹, but in my opinion³⁰, it is the standard view which needs serious revision rather than Lloyd’s hypothesis.

I think that another factor in the Greeks’ unusual independence of mind is their slaveholding. Freedom, in all its aspects, was a highly cherished possession amongst the free, and acute awareness that it could be lost made it all the more precious. Hence their passion for autonomy. Autonomy does not sit well with the notion of a higher authority of any kind, political or intellectual. The atomists (beginning with Leukippos and Demokritos) pushed this to the extreme in side-lining even the authority of the gods, by thinking up a wholly materialistic, deterministic cosmos which had no place for gods at all. The independence theme continues even in Roman times: Seneca, for example, comments with reference to various leaders of and ideas within Stoicism ‘we are not ruled by a king. Each is his own man’ (Epistle 33.9).

All ancient theories on nature were part and parcel of theories on god, the good, and much else besides. The ancient emphasis on the good, the perfect, the form, or similar notions might be compared with the modern hankering after the god of quantification: in both cases there is a tendency to try to import into everything something highly valued in the society which produces it. For some Greeks, such as the elderly Sokrates, what really mattered was how one lived one’s life, and all other matters were subordinated to ethics. Most of us moderns have been brought to believe that what really matters is measurement and accountability (in many senses of the word, but especially the literal one), and all other matters are subordinated to number, even intrinsically incommensurable things such as the artistic element in a performance of ice-skating. For us numbers seem to lend an aura of objectivity, solidity and reliability which mere opinion lacks. But if you allocate a number to a weightless opinion, then you can amass opinions, and get a supposedly objective number; thus the world is quantified. For the ancients, the virtues (especially goodness and justice) seem to have had a role similar to our numbers, providing an idealized scale against which all real things could be measured, a conceptual rock in a real sea of shifting sand, a familiar comfort-blanket in a world in flux.

Ancient atomism had no place for deities and argued for atheism. Competing theories required gods somewhere, e.g. creator (Plato) or unmoved mover (Aristotle), but otherwise tried to keep them out of explanations for natural phenomena. Atomists argued for chaos and randomness as the fundamental principle at work in nature, others (teleologists) argued for purpose and direction, and still others (sceptics) argued for a suspension of judgement on the grounds that the senses are unreliable and, despite everybody’s best efforts, nobody really knows what is or why³¹.

What they all assume, except the sceptics who assume only that they can know nothing, is that natural phenomena occur in regular, ordered ways which are susceptible to analysis. Patterns can be recognized, studied, and explained by reasoning, preferably working from observation of particulars to first principles and then from those first principles to the general case. The imperative for a theory to fit with observation, and the theory’s basis in observation, is clear from the methodological aim to ‘save the phenomena’³².

Pattern recognition seems to have started with celestial phenomena, which are the most regular, if complex, motions in all of nature, and are observable everywhere. Hesiod, for example, is full of folklore on the risings and settings of certain stars and what agricultural jobs should be started or finished by them, and the astral significance of certain days in the calendar. By Aristotle’s time, patterns of an altogether more down to earth form were being sought, and he attempted to sort e.g. the animal kingdom into coherent subsets. Dioscorides sought patterns in the therapeutic effects of different medicines³³, and Galen sought patterns in the structure and function of e.g. various tubes and pipe-like tissues in the body. In such cases the investigator made more or less effort to collect information to analyze. Aristotle’s systematic approach is evidenced better in the area of political science than natural science (but this serves to remind us of the polymathic nature of ancient philosophy and philosophers). He tasked students to find out about the different political constitutions and constitutional histories of well over 100 different poleis, as a preliminary to writing his synthesis of the subject, the Politics. It seems natural to assume that he followed a similar procedure as a preliminary to his syntheses of other subjects, such as the History of Animals.

Another very important aspect of the scientific enterprise in Greece is that audiences needed to be persuaded. They did not – sometimes could not, for reasons just mentioned – judge the validity of a scientific claim on content alone, and they were not in the habit of accepting things on authority, for which the Greeks had a healthy disrespect. So scientists, like politicians, needed to persuade people that they were right and their opponents were wrong. This is true even in mathematics, which also is ‘an art which secures its effects through speech’ according to Plato (Gorgias 451a-c). Style of writing is still today an important factor in the reception of ideas in scientific discourse between scientists, never mind in conversation or communication between scientists and the general public, for example in a court of law dealing with forensic evidence, or media coverage of a scientific breakthrough.

In choosing between rival hypotheses, the best scientists showed a marked desire for simplicity: what is usually called Ockham’s razor ought to be called Aristotle’s razor, according to Derkse³⁸. Simplicity in a theory is not the same thing as a simple theory, or simplification of the problem. As Derkse says (p. 187), ‘it often needs genius (in some degree) to find the simple and convincing solution of what before was only dimly understood’, and ‘the evident simplicity of a solution and its elegant derivation seem to be impossible to overlook only in retrospect’ (emphasis added).

3.    On science and technology and science and philosophy

In modern times there is a clear distinction between science and technology. Science has occupied an elevated intellectual position in recent history, with technology seen very much as a poor relation; science is pure, technology is applied; science is thinking, technology is doing; science is academic, technology is vocational; and most recently in an age of gender-awareness, science is male, technology is female³⁹.

Now the ancient Greeks distinguished between logos and techne, and until recently it has been commonly assumed, without much ado, that logos = reason or word and techne = technology or deed. We might add the parallel assumption that the Greeks were the thinkers of antiquity, and the Romans the do-ers. But even a brief glance at some of the ancients’ surviving texts is enough to raise an eyebrow to this quick and dirty equivalence assumption. For example, in the bible of mathematical astronomy, the Almagest, Ptolemy tells his audience how to make various pieces of more or less complicated equipment for sighting celestial phenomena. Meanwhile in the handbook of the Roman building trade, De Architectura, Vitruvius discusses the relevance of an Archimedean theorem on the sphericity of water to the functioning of a water-level used to ensure that a floor, stair, lintel or whatever was horizontal. The Stoic polymath Poseidonius bothered to describe the nuts and bolts of e.g. weaving, ploughing and baking⁴⁰. In another field, a renowned physician (Andreas, court healer to Ptolemy IV) devised a machine for reducing fractures and dislocations⁴¹. It is relevant here to remember that the doyennes of modern ‘hard’ science, e.g. particle physicists, can be and sometimes must be dab hands at wielding their hi-tech equivalent of saws and screwdrivers, for experimental research on innovative ideas sometimes demands innovative equipment. Innovative equipment, by definition, cannot be bought ready-made, so they must and do make it themselves. By contrast, as Lloyd pointed out (1996b p. 44), the ancient scientists sometimes emphasized the superiority of theory over practice ‘to make a virtue out of a necessity’ because ‘their work often had little chance of practical implementation’.

As Whitney 1990 showed well, technological knowledge was viewed by the ancients as part of a – one – spectrum of knowledge, which ran from the humblest craft activity to the speculations of astronomers. The modern dichotomy between ‘science’ and ‘technology’ did not exist, and the different types of knowledge on this spectrum were classified differently by different philosophers. For example, Plato ordered and ranked them on the basis of their perceived contribution to the development of the moral good of the person pursuing them. This did not stop him, however, from commenting that the proper astronomer would admire the night sky ‘as a geometer might admire diagrams exquisitely drawn by some consummate artist like Daedalus’ (Republic 529d-e). He stops short of saying that Daedalus does geometry, but this needs to be taken in his context that the ‘proper astronomer’ will use the visible heavens only as a guide to aid in the study of the abstract perfect reality. Aristotle produced a less value-laden, broadly tripartite division, consisting of (i) theoretical knowledge, concerned with being and truth (physics, maths, metaphysics), (ii) practical knowledge, concerned with human action (politics, economics, ethics), and (iii) productive knowledge, concerned with making something (house, poem, shield)⁴².

Seneca, disputing Poseidonius’ attribution of the invention of mechanical tools and arts and crafts generally to philosophers, differentiates between sapientia and sagacitas, wisdom and ingenuity⁴³, personified for him by Diogenes the dog (Cynic) and Daedalus (§14). Cicero earlier had linked man’s ability to change his environment through technology with human dignity, a very important Roman virtue, and power, an even more important concept for the Romans⁴⁴. And Augustine later saw technology as an expression of man’s natural genius, even in those cases where its purpose ‘may seem superfluous, perilous and pernicious’⁴⁵. This echoes Pliny’s constant complaints about the vast expenditure of effort on the acquisition or production of such superfluous, perilous and pernicious things as gold and pearls, whilst simultaneously being filled with admiration and wonderment at what people could do.

As regards science and philosophy, the Greeks’ concern with epistemology, with thinking about how they were thinking about a subject, as well as the subject itself, is one of the hallmarks of their work (and is what Crombie emphasises). Nevertheless, the two are distinct, and as there exist many excellent works by philosophers which focus exclusively on the philosophical aspects, so may we focus here on the science. Ideally, the two aspects would be fully integrated, but this is not the place to attempt it. Similarly, cosmology is philosophy, and is be omitted from the pages on astronomy ⁴⁶.

4.    On why it is said that the Greeks ‘invented’ science

In short, because they introduced the notions of natural causality and rational proof: because they tried to eliminate what they considered to be supernatural elements from their explanations for natural phenomena, because they made (often unobserved and sometimes unobservable) connections between phenomena, and ordered them into parts and wholes or causes and effects (rather than just amassed observations), and because they tried to think their way rationally (which does not mean logically or sensibly to modern tastes⁴⁷) through the perceived order of observed phenomena⁴⁸. The belief in natural causation with consequent natural effects was matched by the belief that knowledge proceeds by reasoning from intellectual premise to rational conclusion.

For some reason⁴⁹ a few Greeks living on the margins of the Greek world (and thus simultaneously on the margins of other worlds) began to think that natural phenomena, such as the seasons, eclipses, rainfall and so on were physical consequences or effects of other physical phenomena or causes. Bit by bit, Zeus was relieved of thunderbolt duty, Poseidon of earthquakes, Apollo of epidemic disease, Hera of births, and the rest of the pantheon of gods were pensioned off or abstracted, to become symbols of natural phenomena rather than the willful and larger-than-life but all-too-human characters which populate and sometimes dominate Greek literature⁵⁰. The cause-effect relationships were believed to follow one another in a regular, comprehensible, and (if one’s theories were right) predictable way.

Experience provided observations of particular phenomena, and induction provided a staircase to rise from cases of particulars to higher level generalities. This was the Aristotelian method⁵¹, and was followed on the whole by the Peripatetics, or Pedestrians⁵². The Platonists and Neo-platonists, however, gave more weight to some of the problems connected with sense-perception and preferred not to rely on the senses; they just thought their way through to generalities, sometimes erecting their hypothetical staircase on an implausibly small empirical rock. Many Platonists, as a result, are not often counted amongst ancient scientists, but are labelled philosophers and metaphysicians.

The Greeks did not themselves claim to be doing anything significantly different from what their predecessors did and contemporaries in other cultures were doing. They did not claim to have started a revolution in history: on the contrary, they sometimes even invented traditions which gave the credit for intellectual developments to older cultures, especially Egypt or Babylon. For example, Herodotos credited the Egyptians with the invention of geometry, the Sages of Greece were said to have gone to seek wisdom in far-off lands (usually Egypt and the Near East), and it is commonly observed in modern works that presocratic philosophy/science began in Ionia, the Hellenised coast of modern day Turkey, variously in close contact or conflict with the great cultures of the Near East: Assyria, Lydia, Media, Persia, and Mesopotamia. However, as Jones (1996 p. 154) observed in the context of astronomy: ‘I do not think that the competition [between Ptolemy’s tables and Babylonian arithmetical schemes] would have been seen at the time as one between ‘Babylonian’ and ‘Greek’ science; it is even far from obvious whether Ptolemy and his contemporaries had as clear a notion of the separate Greek and Mesopotamian components in their astronomy as we think we have. At any rate, Ptolemy never speaks in national or linguistic terms, but only of sound or unsound deductive methodology.’

The mathematical fact that we call Pythagoras’ theorem was known to the Babylonians at least 1200 years before Pythagoras was born⁵³. Why then do we call it Pythagoras’ theorem? Because facts and theorems are different. The mathematical fact antedates him, but the theorem is his. And herein lies the difference between Greek activities, which we are calling science, and what went before or elsewhere, which we do not call science. The Babylonians observed a mathematical regularity, and compiled or calculated tables of similar regularities. The Greeks, or rather (as later tradition asserts) a Greek called Pythagoras, observed this mathematical regularity, and proved geometrically that it holds for all particular cases.

The search for the general, the abstract, case, and the process of arguing rationally about the case: these are the hallmarks of science, which are absent from non-scientific knowledge and understanding.

5.    On great men and others in the history of science

There is a strong and understandable tendency in the history of anything to simplify the story in order to clarify it. This applies to academic histories as well as to the popular consciousness. We tend to remember who was first to do something, but not who was second, and we mark some small changes in thought or equipment as ‘innovations’ whilst others are ‘improvements’. Typically the labels are and can be applied only in hindsight. Normally, no-one can tell at the time what ideas or inventions will be significant, hugely significant, or insignificant, to future generations and hence to history. And unfortunately, in the history of science as in other things, the best person does not always win; the fame (very rarely is there any fortune involved) sometimes goes to the ‘wrong’ person or thing. More commonly, the ‘winner’ gets all the fame, and the people on whose shoulders’ he (or very rarely in the ancient world, she) stood get forgotten, and become members of an army of anonymous, industrious, intelligent, nobodies.

These nobodies were an essential element in the story, even if, as in most cases for the ancient period, they are anonymous and will always be so, because history, that is our sources, have omitted⁵⁴ or never knew their names, and exactly what they did. The same problem faces constitutional historians, when they recognise that all sorts of laws attributed to, say, Solon, were not and in some cases could not have been proposed by him, but must have been introduced by some anonymous other man⁵⁵ (and in the political field, it is always ‘man’). It seems to be part of human nature to attribute great deeds to great men, and great deeds fly through time and space like moths to a candle, a great man. But that is not how science or technology develops. It is not a series of great leaps with a few small steps in between. It is a swarming mass of small steps, with an occasional and often unintentional hop, skip, jump or trip along the ways (for there are many)⁵⁶.

Heroes seem to be necessary to all cultures. Certain individuals are widely perceived – rightly or wrongly – to stand head and shoulders above the rest on some culturally determined scale of things, such as, in modern times, kicking a football around, making music, driving fast, or helping others. Different subcultures have different heroes, and heroes exist at all levels, from the school playground to the world stage, past and present. Science is no different. Indeed, modern scientists seem to feel an acute need to indulge in this game, privately and publicly creating and keeping ‘top ten’ lists by subject, by papers, by citation, by era, by anyway they can think of to identify and rank the heroes of their profession.

The ancient Greeks liked to create heroes too. For them, it was typically the top seven rather than the top ten (the symbolism of 7 being profound in Greek, Roman and Hebrew cultures): the 7 Sages, on the science side, and the 7 Wonders, on the technology side⁵⁷. The story is easier to tell when there are heroes. As Homeric battles are apparently won by a few outstanding warriors, so the development of human understanding of and ability to manipulate the world was apparently dominated by a few intellectual and mechanical geniuses. By contrast, a rich description of the story including all known figures tends to be hard for audiences to follow. But in truth, science and technology do not advance by a series of great leaps forward any more than battles are won by duals between champions.

Besides heroisation, the Greeks also liked to make links between individual heroes, to spin continuous threads out of discrete strands, to join the dots to make a coherent picture. The Seven Sages were roughly contemporary and were all in at the beginning of the Greek Renaissance, so they were made into friends and acquaintances. For the rest, in the same way that the gods were united in a genealogy, clever men were linked in an educational lineage, master A taught pupil B, who became a master and taught pupil C, and so it goes on down through the generations. Sometimes such lines of influence have a basis in history, such as Aristotle-Theophrastos-Strato, though even here it is misleading. Theophrastos was a contemporary and close friend of Aristotle; their books were not distinguished by one author or another until centuries later, and Theophrastos was a successor only in so far as he outlived Aristotle by about a generation⁵⁸. Sometimes lines of purported educational lineage indicate nothing more than a similarity in outlook between the predecessor and his alleged pupil.

6.    On hindsight and teleology in the history of science

Cohen and Drabkin chose the best from antiquity, where ‘best’ meant nearest to modern (1940s) ways of thinking. Here we find the ‘greatest’ achievements of Greek science, where ‘great’ is again defined by reference to us, and what we think is ‘right’, or at least, on the right track. This is what some people want from a history of Greek science, and there is a place for it in the modern world, especially for modern practitioners of some science who are interested in the history of their own particular subject. But it gives a very distorted view of Greek science to those who are interested in Greek history and culture, or Greek science as a whole. For example, despite its profound influence on all sorts of topics from alchemy to zoology, four-elements theory as developed by Empedokles and Aristotle is omitted from the contents because it is ‘wrong’⁶¹. As a sourcebook, it naturally omits context. It focuses tightly on the ‘significant’ scientific idea – significant to us, that is. It gives little or no clue of how much else is in the treatise in question, and typically edits out any ‘irrelevant’ material within the crucial passage with a series of … (dot-dot-dots)⁶².

And yet those ‘irrelevancies’ are fundamental for reaching a proper understanding of Greek science. For example, many Greek ‘scientific’ treatises could be used as exemplars in a modern course on rhetoric; a wide range of rhetorical devices are used liberally (in most disciplines), because the author has to persuade his audience (see above §1.2). All this is lost when texts are epitomised or gutted and stuffed into sourcebooks. On the other hand, few of us have the time required to read all the texts from cover to cover, and some subjects interest us more than others. So sourcebooks (like surveys) are starting points, giving quick glimpses of a broad range of subjects. The reader can then decide which subject(s) to pursue in more depth.

Notes

1. According to Crombie 1994 there are, and historically have been, six and only six distinct styles of scientific thinking. A scientific style is a way of thinking about the world that aims ‘to advance knowledge by the identification of answerable questions and soluble problems, to devise methods of finding possible answers and solutions, and at the same time to determine what counted among these as acceptable’ (vol 1 p. ix). Each style focused its inquiries upon certain regular natural phenomena, decided what sort of questions would be considered valid, and determined what sort of answers would be acceptable. ‘A style thus opened certain routes of inquiry and closed others’ (p. xi). The Greek style of scientific thinking, the earliest European style and a style which has continued to the present day, he calls Postulation, and discusses at length in volume 1.

2. Astrology was also closely linked with medicine, chemistry (under the ancient label phusis, physics), botany and anthropology. For the first, see Scarborough 1991 esp. pp. 154-163.

3. E.g. Cohen & Drabkin 1948 p. viii-ix: ‘No one can well deny that a good deal of what may be called ‘pseudo-science’, such as astrology and the like, can be found in the writings of such sober Greek scientists as Aristotle and Ptolemy. But it is well to remember that the intrusion of the occult can be found in modern writings such as Kepler’s or Newton’s and in the contributions to the early volumes of the transactions of the Royal Society down to the works of Lodge, Carrel and Eddington in our own day.’

4. See e.g. Keyser 1990a.

5. See Owens 1991.

6. This is a problem with French’s otherwise excellent Ancient Natural History, which is sometimes frustrating, but rewards well the diligent reader.

7. Knorr 1991 p. 122 n. 11.

8. Needham suggested that this abstract approach arose because ‘the social background of Hellenistic science and technology can be taken for granted because it is quite familiar to us from our schooldays onwards’ 1962 § 26 p. xxvi. This familiarity might have been true for some 30 years ago, but it is sadly untrue today, and a generation of scholars who allegedly took social knowledge for granted, and thus neglected to draw out its significance, seems to have led to scholarship which does not recognise its significance at all, typified in e.g. Barnes 1979.

9. There was a publicly owned gymnasium there from at least the late fifth century, for public use.

10. Any Athenian citizen who wished.

11. See Lynch 1972. The same general point is true of Plato’s Academy; see Cherniss 1945.

12. Which prevented him from committing suicide by mouth after his capture by the Romans, and caused him to fall on his sword instead.

13. We have one name, Krateuas. Very little is known about him, other than that he was known as a rootcutter, he lived at Mithridates’ court, and he wrote a lost, illustrated, work on plants, which may have been the basis for later herbals such as Dioscorides; see Wellmann 1897, with Riddle’s discussion 1985 pp. 180-217. Attalus III of Pergamum was also (and a generation earlier) interested in and very knowledgeable about toxic substances and antidotes, which he tested on condemned criminals (Galen Antidotes 1.1 (14.2 K)), and on ‘friends’ according to the hostile tradition in Justin’s Epitome of Trogus’ Philippic History 36.4.3.

14. This observation was made first by the Greeks themselves. Diodorus Siculus 2.29.6: ‘a few [Greeks] here and there really strive for the higher studies and continue in the pursuit of them as a profit-making business, and these are always trying to make innovations in connection with the most important doctrines instead of following in the path of their predecessors. The result of this is that the barbarians, by sticking to the same things always, keep a firm hold on every detail, while the Greeks, on the other hand, aiming at the profit to be made out of the business, keep founding new schools and, wrangling with each other over the most important matters of speculation, bring it about that their pupils hold conflicting views, and that their minds, vacillating throughout their lives and unable to believe anything at all with firm conviction, simply wander in confusion. It is at any rate true that, if a man were to examine carefully the most famous schools of the philosophers, he would find them differing from one another to the uttermost degree and maintaining opposite opinions regarding the most fundamental tenets’ (trans. Oldfather). There are examples from the medical field and discussion of this point in Lloyd 1995 and Hankinson 1995.

15. Aiskhulos famously wanted recorded on his gravestone not that he wrote this or that tragedy, but that he fought at Marathon. As a hoplite, it is a safe assumption he was a farmer, as were more than 90% of the population. Thoukudides wrote history when forced into exile after active (but unsuccessful, hence exile) service as general in the Peloponnesian war; Xenophon was a professional soldier, again largely because he was exiled and thus prevented from engaging in farming and politics; Cicero was a professional politician first, and author of diverse subjects second; Plutarch was a politician, priest at Delphi and much else besides; the list could be as long as the authors about whom we have biographical information. Why then should we suppose that those who wrote what we call ‘scientific treatises’, about whom we usually have little if any reliable biographical information, be fundamentally different, a class apart from authors of all other types of literary work? For example, later Greeks knew so little about Euclid that they confused the Euclid of Elements with another philosopher called Euclid who came from Megara. See also Authier 1995 on Plutarch as creator of the (false) archetypical image of the scientist, and its perseverance in the face of any and all evidence to the contrary.

16. Nutton assumes the same of ‘most doctors resident in a small town’ and points out that ‘doctors were regularly encouraged to grow their own simples’ [simple drugs], 1985b quotes from p. 140. All those Greek scientists who played a part in the politics of their local poleis (e.g. Archytas, Empedokles, Eudoxus, Hippias, Philolaus) may confidently be assumed to have owned farms. Polybios (3.59.3-5) thought that many Greeks of the Hellenistic and Roman periods pursued an intellectual life because political domination by Macedon, the successor states, and then Rome, ‘relieved’ them of the ambition to pursue a life in war and politics when not managing their farms. Polybios himself was a case in point.

17. By Saïd al Andalusi in the Book of Categories of Nations1991 p. 26. The association is not accidental: the geometer’s tools are the carpenter’s tools – compass and set square or gnomon. Hahn (1995 p. 126 n.25) drew attention to ‘a kind of applied geometry with technological innovation’ which characterizes the real or attributed achievements of Thales, Anaximander, Rhoikos, Theodorus, Khersiphron and Metagenes. Although only the first two are generally thought of as presocratic philosophers (the other four being labelled architect/engineers) I believe that the kind of employment specialisation these labels imply is inappropriate for the period.

18. Namely, Celer the centurion, Diogas the trainer, Euschemus the eunuch, Flavius the boxer, Orion the groom (which indicates an overlap between human and veterinary medicine), and Philoxenus the schoolmaster, all cited by Nutton in his hugely entertaining, as well as very informative, article on ‘The drug trade in antiquity’ 1985b p. 145.

19. Cicero, nearest in time of those remarking on his origins, described Arkhimedes as ‘a humble little man’ Tusc.Disp. 5.23, which can hardly refer to his personality as revealed in his surviving works, and therefore ought to be read as a reference to his socio-economic status. Silius Italicus thought likewise, calling Archimedes destitute (nudus) Punica 14.343. This, however, did not fit at all with Plutarch’s image of the scientist induced only by Roman soldiers besieging the city to tear himself away from contemplation of abstract mathematics and get his hands dirty in an occasional bit of mechanical engineering, so Plutarch Marcellus 14.5 says he was a relative (unspecified) and friend of Hieron.

20. Aristea Epistle 10, 29-30. He too was ‘not well-born’ according to Diog. Leart. 5.75.

21. E.g. Poseidonios of Apameia and famously Karneades, who so irritated Cato (Plut. Cato 22-3). Similarly the historian Polybios, who went to Rome as a political hostage for the good behaviour of his home community, the Achaian League, after they lost the battle of Pydna.

22. His treatise on the Aqueducts was completed under Trajan. Before that he had written military treatises for Domition, and a work on surveying.

23. Absence of evidence is not evidence of absence, and the primary sources (prior to the C2 AD) are notoriously uninterested in biography and autobiography, so we must be content with probabilities. Greek scientists did not leave memoirs – even Galen did not leave a memoir as such. On his many autobiographical comments scattered through his works, it is essential to understand the context, for which see Nutton 1972, to whose points we should add one more, concerning forgeries. ‘Forgers’ might not simply be in the business of trying to make sales of their own works by passing them off as Galen’s; they might rather be medical opponents in the business of trying to discredit him, as Anaximenes of Lampsakos ‘forged’ one of his rivals, the historian Theopompos, to damage him (Pausanius 6.18.2). There are of course secondary sources written in antiquity, by men like Diogenes Laertius (Lives of the famous philosophers), Philostratos (Lives of the sophists) and Eunapios (Lives of the philosophers), but they are all late (C3-4 AD) and generally unreliable – they are mentioned frequently in modern scholarship because we are beggars for biographical information and have no other choice. On the hypothetical Greek boy, my feeling is that the more ‘elite’ the father considered himself and his family to be, the more opposed he would have been to this idea.

24. Much of what he has written since 1979; see e.g. Lloyd 1992. See further below §2.

25. For examples illustrating the importance of this type of approach to reaching a fuller and deeper understanding of the scientific content of such texts, see e.g. Klein 1968 on mathematics (arithmos means a number of things, not just an abstract quantity); Osborne on the importance of the literary context in which fragments of the pre-socratics appear in later authors, Rethinking early Greek philosophy; Solmsen 1975 on pre-socratic and socratic ways of thinking evidenced in history and drama (and other forms of non-philosophical/scientific literature); Rihll and Tucker 1999 on the socio-economic context of theories of matter; Cosens 1998 on the importance of practicals in Galen’s treatises; and Shapiro 1994 on seventeenth century scientists’ adoption of practising artists’ ideas on colour mixing (this has much interesting material on ancient theories of colour).

26. Introductory essay to the reprint of his Inaugural Lecture, in 1989 p. 353.

27. For example, Zeno of Sidon launched what he thought was a damaging attack on Euclid’s axiomatic method. Poseidonios (amongst others) argued the toss with him; see Proclus On Euclid’s Elements Book 1 especially 199-200, 214-218.

28. Which originated with the French scholars Gernet, Vernant, and Detienne, who argued for an association between rationality and the polis, but which has been developed by Lloyd.

29. This envisages an ‘aristocratic’ stage in the (crucial) archaic period, with a very gradual assumption of responsibility by the demos (the people), and presumes that what debates were held in this period were dominated by a few ‘great and good’ whilst the demos stood around and said or did little or nothing. This is hardly compatible with the general atmosphere of argument and debate presumed by Lloyd. He has been criticised for this reason in Hurwit 1985. The details needed to try to dovetail social and intellectual histories are also unexplicated at the moment: since the intellectual revolution started in Ionia, does Lloyd suppose that the Ionian poleis were more ‘democratic’ earlier than the mainland poleis? Is there any correlation between constitutional type and philosophers’ home towns? Although Athens is the most famous and most well known ancient democracy, her only homegrown critical thinkers of great stature were Sokrates and Plato. Many people from other poleis found this polis a congenial place to live and think critically – at least some of the time (Anaxagoras and Aristotle, amongst others, both felt compelled by fear to skip the country at a certain point in their residencies). Which conjunction leads to the question: is there a connection between being a metic, rather than a citizen, and being a free thinker? Metics, by definition, were excluded from the political environment, so if there is a connection here, it would considerably complicate Lloyd’s hypothesis. It would however tally with Polybios’ view on political and intellectual life (see n.16 above) .

30. I speak not as an historian of science but as an ‘ordinary’ ancient historian who cut her professional teeth on Greek political and constitutional history, especially in the archaic period.

31. See Sharples 1996 for a discussion of the main schools in Hellenistic times, arranged around their answers to various types of question rather than a traditional historical narrative or description of each school’s philosophical tenets.

32. On which see Lloyd 1991 chapter 11.

33. See Riddle 1985.

34. After a couple of hundred years of such data gathering and processing it became possible for others to write the first encyclopaedias, which suited the Roman temperament of the times.

35. The ‘museum’ at Alexandria was not a museum in the modern sense, but a home of the muses, that is, a place where the muses’ arts were practiced. The main facilities were the library and a zoo. Little zoological work of consequence was done there and I believe that it functioned more for the entertainment and adornment of Ptolemy’s court than for any investigative study of animal life. Pickled specimens are recorded in e.g. Demostratos apud Aelian 13.21, but these are oddities or curiosities – in this example it is a supposed Triton, half man-half fish, preserved at the inland town of Tanagra.

36. Even on the most pro-active hypothesis about how Theophrastos obtained his information (Maxwell-Stuart 1996), it is supposed only that he ‘took notes of the places through which he passed in the course of his everyday life, as events directed’ (p. 266). There is no suggestion that he travelled to find answers. He did, however, start a garden (according to Diogenes Laertius’s life – written over 400 years after Theophrastos’ floruit) in which he might have grown and studied a number of plants, but he could not possibly have grown the vast majority of the plants mentioned in his botanical works, the information for which surely came from oral, and to much lesser extent written, sources.

37. E.g. Archimedes On plane equilibriums prop. 7 (principle of the lever): ‘If the magnitudes be incommensurable, they will likewise balance at distances reciprocally proportional to their magnitudes. Let (A+B) and C be incommensurable magnitudes, and let DE, EF be distances, and let the ratio of (A+B) to C be the same as the ratio of ED to EF; then I say that the centre of gravity of the magnitude composed of (A+B) and C is E’ (Thomas trans. slightly modified).

38. Derkse 1993 p. 203. ‘Ockham’s razor’ is the methodological view that in the formulation of explanations, ‘it is better to use the most limited set of explanatory elements (hypotheses, assumptions, variables) wherever possible and if adequate. Superfluous explanatory elements should be shaved away’ Derkse 1993 p. 10, emphases added. When choosing between competing explanations the same principle can be followed. Derkse argues strongly that Aristotle uses simplicity as a constant concomitant in the inductive process of intuitively grasping first principles, and that it could itself be labelled a principle on Aristotle’s theory of knowledge. ‘[Aristotle’s] abundant use of some form of the principle of parsimony, economy and simplicity is to be found in many domains and has many characteristics: as a principle of minimal ontological assumption (Physics), as a rule of method (Post. Anal.), as a criterion for theory evaluation (On the Heavens, Physics), as a heuristic device (in the biological writings), as a surprising feature of the workings of nature, which gives aesthetic satisfaction and intellectual joy (biological works)’ p. 203.

39. See e.g. Cowan 1996 or Wertheim 1997.

40. Seneca Epistles 90.20-23. Poseidonius’ attention to ‘banausic’ arts may have disappointed Seneca, but Poseidonus clearly thought these subjects were worthy of serious intellectual effort.

41. Celsus 8.20.4. See also Drachmann 1963 pp.171-85 with translations of relevant parts of Rufus, Heliodoros and Galen apud Oribasios, citing devices invented or improved upon by a number of named individuals who by implication were doctors. For in 49.23, when Oribasios mentions the trispastos invented by Apellis or Archimedes, he distinguishes them by saying ‘First let us bear in mind that neither Apellis nor Archimedes was a doctor, but they were mechanics’. He continues that ‘the doctors of that time reduced the dimensions of the construction and made out of the triple pulley a surgical spanner for resetting dislocations and fractures’. Other examples of the point: Knorr 1991 argues that the Elements has a strong basis in practical applications, and Høyrup 1997 argues that (i) ‘Hero’s geometry depends to a greater extent than [is] usually assumed on Near Eastern practical geometry or its descendant traditions in the classical world’; (ii) that the conventional image of Heron ‘as the transformer of theoretical into applied mathematics is only a half-truth’; and (iii) that ‘much of what is shared by Hero’s Metrica and the pseudo-Heronian collections assembled by Heiberg as Geometrica are shared borrowings from the same [practical geometry] tradition’ (quotes from p. 67). O’Neill 1998 argues more generally that the distinction between ‘practical’ (problem solving, constructing) and ‘theoretical’ (seeking after truth, discovering) mathematics is overplayed and not helpful.

42. Whitney discusses these points fully, with ample references to the primary sources, in chapter 2.

43. Epistle 90.7-13, esp. 11.

44. De natura deorum 2.50. Cicero follows his teacher Poseidonius in crediting philosophers with the creation of civilisation, even if he disagreed with him on much else; see e.g. Tusc. Disputations 5.2.

45. City of God 22.24. Whitney thinks that Augustine is being ironic in this passage. Identifying irony in writing is always a contentious issue, but in the area of scientific texts I agree with Scarborough and Nutton 1982 p. 214 that ‘irony is always the last refuge of the baffled translator’.

46. The distinction between and mutual independence of cosmology and astronomy in the ancients is clarified and emphasised by Hanson 1973.

47. See Frede and Striker 1996.

48. Lloyd 1983 emphasised that ‘scientific’ writings (particularly the life sciences, on which he was focused) sometimes do little more than repeat folklore, but then there is an attempt to rationalize those traditional beliefs, which sometimes involves examining them critically and testing them. He also stresses the difference between what the scientists say they should do or are doing and what they actually do do, which usually falls short of these ideals to greater or lesser extent.

49. Many hypotheses have been advanced to explain this. Farrington 1944 and 1947 emphasized the role of technology. A contributory role for technology has also and recently been argued by Hahn 1995. I concur fully with his view of the influence of then current technology on ancient philosophers, at least in the Greek period – before the sort of attitude exhibited by Plutarch set in, and which Plutarch anachronistically attributed to earlier men about whom he wrote, e.g. Archimedes (Life of Marcellus) or Plato (Convivial Questions 8.2.1). Goody and Watt 1968 pp. 27-68 emphasized the role of literacy. Most recent views take Lloyd’s socio-political context as a starting point. Many add what might be termed cultural hybrid vigour (Sarton argued this long ago): those living in the Ionian communities, exposed to more than one cultural influence (Greek and Lydian/Median/Persian), were presented with radically different explanations of the world. I would emphasise that this circumstance was compounded by the developing slave culture which forcibly moved and rapidly assimilated different Greeks and non-Greeks into new societies and cultures. As such, faced with a choice of different, competing, traditional, ‘received wisdoms’, they were prompted to sort and criticize them, from whatever quarter they came.

50. See e.g. the discussion about Jupiter in Seneca Nat Quest. 2.42-46.

51. See e.g. Topics 1.10, Posterior Analytics 1.31.

52. ‘Pedestrians’ is Saïd al-Andalusi’s translation of the term peripatetic, Categories of Nations p. 29.

53. See Aaboe 1964.

54. It is quite noticeable how often ancient critics of axial rotation or heliocentric theory omit the names of Herakleides, Hiketas, Aristarchus and Seleucas, and refer anonymously to ‘some people’ or similar such terms. Anonymity does not just effect the ‘little people’.

55. For example, the story that Solon introduced laws ‘reforming’ the coinage. Athens almost certainly did not adopt coinage until after Solon’s floruit.

56. See e.g. Boas 1959 p. 499 on Lavoisier and modern chemistry: ‘As usually happens when one examines a dramatic event in the history of science, it turns out not to have been so simple. Seventeenth-century chemists had tried to have both a rational system of nomenclature and a rational theory of combustion; had tried, and had failed’. Or Ihde’s comments on the same paper in the same book, p. 522, ‘this period [was] one which was essential for the clearing out of a great deal of rubbish which had accumulated during the previous centuries and which was serving as an obstacle to progress in chemistry. Actually a study of this period is a confusing one because, as so often happens with the clearing out of rubbish, it is not destroyed but is tossed aside where it can again serve as an obstacle to progress’. The rubbish in question is Empedokles’-Aristotle’s four elements theory, by the way.

57. There are many examples of sevens in Greek thought: see Byl 1980 pp. 252ff for its appearance in Aristotle.

58. Aristotle died first (aged 60-something) whilst Theophrastos enjoyed another 30 years or so of vigorous intellectual life, continuing to study, teach and write into his 90s.

59. Theories ‘that are now known to be false or even ridiculous’ were referred to in notes only as necessary to make sense of a passage quoted. They omitted entirely anything which ‘encroach[ed] on the field of Greek magic, superstition, and religion’ (all quotes from p. viii). Lines have to be drawn somewhere, deciding what to include and what to exclude. We would draw them differently today, and 50 years from now they would be drawn differently again. For example, a contemporary sourcebook might include as (superficially) an apparent antecedent to gene theory Seneca Nat Quest. 3.29.3: ‘In the semen there is contained the entire record of the man to be, and the not-yet-born infant has the laws governing a beard and grey hair. The features of the entire body and its successive phases are there, in a tiny and hidden form’.

60. And the effort is very considerable. See for example Touwaide 1991. About 5% of the collection of Greek veterinary treatises which make up the Hippiatrica have yet to be edited, according to Fischer 1988 p. 195.

61. Of course it becomes necessary to explain what this theory was in order to understand reference to it (actual or implied) in passages which are selected for inclusion in the sourcebook, and Cohen and Drabkin provide such an outline in the form of notes. But they do not include any source specifically on this theory or its main ancient competitor, atomism.

62. In this respect Lewes 1864 stands apart from most scholars for his decision to retain the ‘essential or incidental absurdities’ thus ‘preserving, as far as may be, the historical colouring derived from the inherent weakness of early science and the individual weakness of Aristotle’ (p. ix).

FURTHER READING

A good brief introduction to the subject as a whole is G E R Lloyd’s Early Greek science: Thales to Aristotle and Greek Science: Aristotle and after both Chatto & Windus 1982. R French’s Ancient Natural History (1994) ranges widely across topics and time, organised loosely around the chronological development of ideas about nature and how to approach it. It has a thought-provoking introduction which, in a nutshell, argues that the word science is inappropriate in the ancient context. D Lindberg’s The Beginnings of Western Science (1992) provides a more traditional type of overview. Another wide-ranging work on Greek science is M Clagett Greek science in antiquity (1955) which, despite changes of emphases in the philosophy and history of science during the 45 years since it was written, is still worth reading to get an overview of the subject. Slightly older, but also still valuable, is B Farrington’s Science in antiquity (1936) and Greek Science (1944, 2 vols). T Africa offers (in his own words) ‘not a history of science but a study of scientists’, which concentrates on context, and which foreshadowed many current debates about Greek science (e.g., the competitiveness of ancient scientists and their theories, the continued inclusion of folklore within the ‘scientific’ tradition, and the status of astrology and other practices now considered occult or hokum) back in 1968: Science and the state in Greece and Rome.

More detailed treatments of the nature and beginnings of Greek science can be found in Lloyd’s four books Polarity and analogy (1966) which is concerned above all with method in ancient scientific thought, Magic, reason and experience (1979), Science, folklore and ideology (1983) and Aristotelian explorations (1996), which all have special reference to the life sciences, but the arguments have wider application. A number of his articles on specific topics (again mostly but not exclusively concerned with the life sciences) have been collected and reprinted in convenient form in Methods and Problems in Greek science (1991). This reprint includes very useful introductions which ‘take stock of work done subsequently on the problems discussed’ (p. xiii) and which thus point to the directions in which some of the arguments have moved and the changing emphases which Lloyd and others have put upon each of them. His Adversaries and Authorities (1996) compares and contrasts ancient Greek and ancient Chinese approaches to science, illuminating all sorts of issues concerning the practice and practitioners of science in these societies.

There are a number of useful electronic resources and web sites for the history of science. There are links throughout the site, especially to texts and translations.