www.TheLogician.net© Avi Sion All rights reserved


© Avi Sion, 1990-2005 All rights reserved.

You can BUY online, Amazon.com (in paperback or kindle/.mobi form), at Lulu.com (in hardcover, paperback or e-book / .epub form ), and at many other online stores.


Part II, Chapter 7:(Chapter 13 in 2008 reprint)

Comments on 3 chapters of Foucault

These notes on the first three chapters of Michel Foucault’s The Order of Things[1] were written in 1990, prior to writing the essay Logical Aspects Of Michel Foucault’s “Archeology.My intention was a chapter by chapter commentary of the book, but I did not finish that. The comments are nonetheless interesting.

1. On Chapter 1

2. On Chapter 2

3. On Chapter 3

1. Las Meninas.

Apparently, a rather longwinded demonstration, with reference to a painting by Velazquez, that real things, events and relationships are infinitely complex, and capable of interminable verbal description. Whereas, once proper names are introduced, the inherent pregnancy and polyvalence of the original is effectively abandoned.

But, of course, it need not be so. The names do not in themselves arrest further description; they are exact parameters, but all the ambiguities beyond them are still operative, and still open to discourse. Nor, as Foucault admits, does the ‘infinity of the task’ allow us to infer that ‘words are imperfect.’

2.The Prose of the World.

An interesting analysis of the way the world was thought, until the end of the 16th century, in Western culture, at least according to Foucault. This refers supposedly to the Medieval Christian and Renaissance cultures. He considers that this period shows a distinct epistemological framework, in comparison to the more ancient GraecoRoman cultures.

This new ‘logic’ (let us say) centered on an (to our eyes) extravagant concept of ‘resemblance’ (which later became more refined and stringent, in the ‘Classical Age’ of the 17th century and on). I am not sure of the correctness of this perception: neither that there was a historical discontinuity, nor that ‘resemblance’ was so universally understood by the pre-Classicals in quite the way Foucault posits. His interpretation of events is, to be frank, a bit simplistic.

Foucault: Four main kinds of ‘resemblance’ were claimed. Things may be adjacent in place or time, ‘convenient’ (proximate); they may be mirror-images of each other, ’emulate’ (or imitate) one another; they may be more abstractly and remotely similar, ‘analogous’; or they may be ‘sympathetic,’ one tending to become more like the other-though such change toward identity and singularity is held back by an opposing force of ‘antipathy.’

I say: It is especially the last principle which is at stake. The ideas of Sympathy and Antipathy were perhaps a physical theory (natural science). The ‘infinite reflection of the Object, even into words, or other shapes and sounds,’ idea is tenable, and has some truth. It may be that a large number of writers used this methodology to excess, I am not contesting that. So another interpretation of events, than that offered by Foucault is perfectly feasible.

Even today, the idea of affinity plays some role in our thinking. Does not the butterfly look like the flower? Do not husband and wife often come to strikingly resemble each other over time? The idea of evolution of species by natural selection and adaptation surely contains echoes of this: a process tending to certain uniformities. Modern Chemistry appeals to ideas of attraction and repulsion, to explain chemical compositions.

Foucault: These resemblances were knowable by reference to more or less hidden ‘signs,’ which were themselves resemblances of sorts. Thus, for instance, the medical value of walnuts was suggested by the similarity in shape these fruits have to the human skull and brain. Since the semiological relation was essentially one of similitude, the epistemology of that period was much less restrictive than our own.

This writer: I am not sure that the method of looking for signs (themselves resemblances) can ‘archeologically’-speaking be regarded as the overriding methodology of the period. People were still perceiving, they still had senses, they still conceived abstractions as we do; they still referred to logic, and were already cognizant with the Ancient works on the subject, including the Organon. It would not be accurate to characterize the 16th-century-or-earlier epistemology as exclusively focused on concepts of Sympathetics origin.

Foucault: Anything to do with an object of study, however incidentally or accidentally, was equally significant. Magic and divination were accepted as on the same level as erudition, and ancient texts and commentaries thereon were as relevant as direct observation and independent reasoning. Words about something were part of that thing, and therefore their repetition constituted knowledge, as ‘objective’ as any other.

Self: The facts presented here are quite conceivable, though I tend to be skeptical that they were historically as widespread as Foucault suggests. In any case, he quite rightly points to the Judaic roots of this indiscriminate methodology. But it would be unfair to squarely blame these developments on Biblical beliefs. Extrapolations were made by certain (let us say) Christian thinkers; but these extensions were not conversant with the traditional parameters of applicability, or chose to ignore them. Thus, Jewish logic can in no wise be blamed for these developments, to the extent that they occurred.

It is true that, according to traditional Jewish philology, the world was created through (a primeval version of) Hebrew, which language therefore reflected, in its sounds and shapes, the essences of the things it referred to. However, according to that same tradition, this strong relation between sign and signified has been considerably diluted and distorted since the Babel incident, so that inferences are only possible within very strict limits, known to only a few Sages.

Mediaeval Christian attempts to hang on to a methodology that was no longer so applicable (to the untrained user), and especially not to post-Babel languages (even though they retain reflections of the glory of the Holy Tongue), were therefore unjustified within the framework of the doctrines they claimed to echo. Judaism certainly never intended to foster superstition or ‘blind faith’ in ancient philosophers, quite the contrary. The World was created by God as an act of mercy, of love; it is essentially a benevolent, beautiful place, with a great potential for dignity and decency.

Further on, Foucault suggests that, as of the Renaissance, the written word acquired a predominance over the spoken. That is doubtless true, what with the advent of printing and the spread of literacy. He is attempting to show that the written word made a quantum leap, at about the same time as the Sympathetic methodology reached its peak. Okay, but there was not such a sharp distinction as he is trying to imply. In that case, so what? There were shifts in emphasis, there always are; the overall cognitive process is in essence the same, with in it the seeds of many alternative expressions.

In any case, Foucault’s statement that ‘the Law was entrusted to the Tablets, not to men’s memories’ is inaccurate, from the Jewish point of view. For Jews, the oral transmission was always as weighty as the written one, if not more so (in the sense that those who lack the oral, cannot fully understand the scriptural). The relatively modern concentration on written texts must be viewed rather as resulting from a gradual breakdown in the social cohesion necessary for oral traditions; study became a more individual activity, and therefore one more dependent on the written word.

Lastly, Foucault points out that the assumption that words and things are related by virtue of ‘resembling’ each other, began in the 17th century to be displaced. The relation gradually became more tenuous and arbitrary; though this tendency, he claims, has been somewhat prevented from excesses thanks to modern ‘literature.’ In my view, this ‘symbolization of symbols’ was a positive development in itself, though one which could be misinterpreted, and indeed has been by some.

Although modern language cannot claim an intrinsic power of representation of reality, it still depends for its meaningfulness on perceived, conceived or assumed distinct similarities between the objects it refers to. The sound and shape of the word is arbitrary and what we choose to attach it to is our prerogative; but the word remains meaningless if we are not agreed-upon using it to refer to some objective individual entity, or group of entities with distinctive common factors. Comparison and contrast remains the foundation of conceptual knowledge.

Thus, Professor Foucault is tending to over-generalize; he speaks in flourishes, sans l’extrème rigueur de pensée qu’il se doit d’après ses propres arguments. The relation between words and things, or knowledge and reality, was not understood by the pre-Classicals as simply as Foucault implies. It is not accurate to suggest that ‘resemblance’ (in the indiscriminate sense above described) characterized their episteme so thoroughly; there were other points of reference too. For instance, causality is not mentioned by Foucault in this context; yet Aristotelian influence did exist at the time.


By this term, Foucault formally intends the ‘semiological’ relation between a sign and its object. He claims that this relation had something to do with ‘resemblance’, prior to the 17th century and on; at which time a stricter version of similitude was adopted, the method of discriminate identification. This method, of cautious analysis of identities and differences, as suggested by the likes of Descartes and Hume (though with different emphases), was a new episteme.

Thus, in Foucault’s intent, ‘representing’ is a neutral general term for the sign-object relation, however any period may believe that this relation is specifically established. It simply means ‘taking the place of’ or ‘standing in for’, without implying a re-presentation, in the sense of a similitude, between sign and object.

In any case, Foucault has confused the ‘sign’ in the sense of the walnut’s shape being a sign of its medical qualities, with the ‘sign’ in the sense that a word is a sign for its object. The of/for difference in preposition is important; it makes two words out of the one, the noun ‘sign’. Pre-Classical thinkers may have subscribed to the walnut-interpretation method, but it does not follow that this in any way modified the sign-object relation.

They were just accepting another kind of object, or phenomenon, which we no longer rely on so seriously; the ‘archeologically’-implied relation of signification was unaffected. Foucault himself formally admits the perils of ‘establishing discontinuities’ particularly in the history of thought. Yet, the content of his thesis is replete with such ‘arbitrary division’.

Thus, he characterizes pre-Classical thought as justified by the sum total of its parts, whereas the Classical ‘exhaustive census’ gave rise to ‘absolutely certain knowledge.’ Both of these characterizations are exaggerations. The two periods are not distinguishable with reference to these characteristics, because both of them involved both contextuality and enumeration to some extent.

Astrology and alchemy imply a looser, more poetic methodology; astronomy and chemistry, a more precise and mathematical logic. With regard to the ‘representing’ relation, which others have called more broadly ‘signification,’ it existed prior to science as well as in science. The difference between the two periods is this: the former was not as conscious of identities and differences as the latter; and the latter relied on ‘resemblance’ as much as the former, though in a more cautious and thoughtful manner.

Foucault presents Don Quixote as the first modern character of literature, as well as the last hanger-on to the methodology of consulting texts for an externally suggested world-view and behavior-pattern. He is depicted as formatting his thought and action, in slavish accord with the ideas and examples of ancients, and rejecting as unreal, as magic, any personal insights or perceptions which disagree with his loyalties. He is bound by resemblance to mythical characters and events; he reads nature through books, discarding independent epistemology or conscience.

Thus, Don Quixote suffers from a sort of insanity, an alienation into imagination and analogy. His fiction becomes a reality, when his early adventures are in turn published in book form while he is yet alive. We learn from this that language ‘now possesses new powers.’ In my view, all this is, again, exaggeration. Everyone relies to some extent on received knowledge, from previous generations or others in one’s generation; everyone’s behavior is to some extent influenced by other people, whether in writing, by speech, or by example.

It would therefore be unfair to characterize Don Quixote, or the episteme and period he represents, as peculiarly dependent and bumbling. He seems like a piously fanatic fool to our eyes; but who knows the inner development he was going through. Think for instance of the mediaeval churchman (Frollo, the cathedral’s archdeacon) in Victor Hugo’s Notre Dame de Paris. He is painted much more sympathetically, as a madly impassioned searcher; and he emerges as a credible construct (though ultimately tragic, destructive). Foucault has focused on just one possible characterization.

Furthermore, modern man is no less involved, in his own way. We all have our authorities, our points of reference, our trials and errors, our imaginations, our boundaries. That fact in itself does not disqualify someone; the only issue is how wisely we absorb others’ contributions and handle our finitudes, how consciously and selectively. People vary in intelligence or virtues; some make more mistakes than others. Every period has its achievements, and its limits. But nothing has changed, the epistemological background is the same; we are all to some extent Don Quixotes.

I mean, just look at the power exercised by today’s medianovels, movies and T.V. Their whole raison d’être is producing role-models, and it is no accident that actors are referred to as ‘idols.’ To suggest, as Foucault does, that we have become free of such dependencies is gross inaccuracy. Consider Jean-Paul Sartre’s distress at the difficulty of spontaneity; a modern man, locked in a self-made prison of role-play. Sartre simply replaced traditional models with one of his own fancy; but the art of natural behavior still eluded him.

There is nothing intrinsically wrong in referring to tradition. The ‘wisdom of the ancients’ is neither proven nor disqualified by its antiquity. Many of the speculations of ancient philosophers are still of interest to us today; not merely as historical opinions, but because they continue to enrich and stimulate our thought. ‘Authority’ is often well-earned. Scholarship did not suddenly die; and the proof is afforded by Foucault’s own research into past thought. As Anatole France suggested:

Any expression of an abstract idea can only be an allegory. By an odd fate the very metaphysicians who think to escape the world of appearance are constrained to live perpetually in allegory. A sorry sort of poets, they attack the colors of the ancient fables, and are themselves but collectors of fables. Their output is mythology, an anemic mythology without body or blood. (Bentwich, 345.)

I am not trying to play down differences, but merely to put things in perspective. Let us continue. Foucault says, with reference to the Classical period as of the 17th century: ‘there can be no sign until there exists a known possibility of substitution between two known elements.’ The subject-object relation is brought into the equation between sign and signified. Thus, he in effect perceives a shift from de re modality to the logical, de dicta mode.

Signs were no longer ‘representative,’ in the sense of microcosmic reflections of objects, but more frankly conventional. The ‘resemblance’ factor was relatively diminished in the relation of signification. A sign (read: word), henceforth, contained within itself a statement of its function as a sign, as well as a statement as to what it specifically referred to; but otherwise its relation to the object was man-made.

Fair enough, but I disagree with Foucault’s analysis. The natural causality referred to, when for instance we take a cry as a ‘sign’ of a babythis is still with us, even today; it was not abandoned in the Classical period, and nor was the belief that ‘if no one were to perceive’ things, they would be ‘just as much there.’ Similarly, it would be inaccurate to say that logical modality was absent from pre-Classical thinking processes (witness Maimonides’ critique of the Arab Mu’tazilites school, for instance).

It is true that philosophers like Descartes, Leibniz, Bacon, Berkeley, Locke, Hume, discussed the ‘connections of ideas.’ This was an outcome of their analysis of sensory-perception as a physiological process terminating in the production of mental images, called ideas. They did not see the paradox generated by this hypothesis, that if what we perceive are ideas, then how do we know of an object capable of producing them? It was an erroneous approach, which was only later corrected by a more phenomenalist ordering of events.

But in any case, a distinction must be made between the methods professed by the philosophers of a certain periodtheir own understanding of what was going on in their milieuand the methods actually used by human beings of the time, themselves included. The former belong to the history of explicit philosophy; the latter, more broadly to cultural history. These two processes are not always, if ever, at the same stage of development. Human methodology changes little, shifting in emphasis, but not in its essential components.

For these reasons, it seems to me that Foucault’s suggestion that the semiological relation itself underwent a radical structural change, is rather hyperbolic. The potential for words to serve as ‘transparent and neutral’ symbols coexisted with the more florid view of language as ‘one of the figurations of the world’certainly the former is found in Aristotle, at least. And as for applying ‘one and the same name… indifferently to things that are not of the same nature’it is an error we all still occasionally make.

Now, Foucault offers the following epistemic constructs, as characteristic of the new order. Instead of an emphasis on ‘resemblance’ (to ridiculous extremes), a more pondered observation of differences. Instead of far-fetched and vague analogies, ‘complete enumeration’ of cases and the elements in each case, with a more discriminating eye. Separation of historical and scientific research, so that the opinions of past authors are regarded with a more critical eye, if at all considered; they are no longer authorities, though they may remain contributors.

Science orders information either in the way of a mathesis, with reference to precise measurement of numbers or degrees; or at least in the way of a taxonomia, a more analytical ordering of data with reference to qualitative identities and differences. Additionally, the genesis of things and ideas must be considered; this is the chronological and epistemological aspect of science. The whole has to be empirical, yet imagination is also required to reconstitute an order.

The Rationalist/Empiricist divisions between Classical philosophers, then, reflect different emphases within that framework; but in any case, according to Foucault, both differ radically from the preceding period of ‘divination’ methodologies, which made more comparisons than distinctions and failed to carefully observe the object itself before flying off into romantic associations.

However, Hume’s comment on the pretensions of the new philosophers is apropos:

Let the philosopher pride himself on his precision as much as he will… I nevertheless defy him to make a single step in his progress without the aid of resemblance.

Similarity (moderated by dissimilarity) was always, and continues to be, the basis of all conceptual knowledge. It is possible that the preceding period involved more imagination of resemblance than was justified, but it is impossible for the basic relations to change.

Formally, a word X is related to some pointed-to thing or group of things, by the statement ‘X is a sign for this/these thing(s)’this is how the relation of signification is defined, without any presuppositions as to the particular configurations of the thing(s) referred to, or the basis for their being grouped together. It is true, as Wittgenstein objected, that indication (pointing to something, saying ‘I mean this’) is itself a vague act; but context-changes gradually purify such ideas of possible ambiguities or equivocations.

The label may or may not itself contain other relations (like similarity of sound or shape) to its object; and putting a label on a group of objects does not guarantee that they possess a distinct commonalty other than the arbitrary label itself.

According to Foucault, the relation of signification became a component of the sign, instead of a copula linking sign and object. He claimed that ‘no specific activity of consciousness can ever constitute a signification,’ and inferred that signs changed from ternary organization to a binary one. But this seems a forced, hair-splitting argument to me. It matters little whether we regard the copula as in or out of the sign-term, or its genesis as arbitrary or imposed by some resemblance between sign and signified. There is always a final implicit thought ‘X is to be the sign for the indicated thing(s)’.

I am not at all convinced that the Logique du Port-Royal was introducing a novel sign-object relation. It states: ‘The sign encloses two ideas, one the thing representing, the other the thing represented.’ I do not see this definition as formally excluding the Renaissance interest in what makes possible ‘to see in the first the mark of the second.’ The Renaissance’s specific answer to that question may have been fantasy-prone, but the question in any case remains operative.

Funnily enough, in my view, the Classical philosophers unwittingly created a new problem, by confusing things and ideas. The above Port-Royal definition is a case in point. The sign ‘encloses’ the idea of ‘the thing represented,’ they said; but in fact the sign is supposed to refer to ‘the thing’ itself, not to the idea of the thing. Whether a mental entity called an ‘idea’ stands between the label and its object is an open question. A broad, neutral definition cannot at the outset exclude a direct subject-object relation.

Ideas may exist, as memories of previous perceptual and conceptual acts, without implying that these acts require intermediaries. Ideas may be sometimes formed on the basis of imagined realignments of the mental images of some concrete and/or abstract components of things; but it does not follow that they are always so formed.

I am not, of course, denying the great value of epistemological and philological contributions of the Enlightenment period, but merely to some extent disagreeing with Foucault’s interpretations of these developments.

[1] Foucault, Michel. The Order of Things. An Archeology of the Human Sciences. 1966. New York: Vintage, 1973.

You can purchase a paper copy of this book Books by Avi Sion in The Logician Bookstore at The Logician’s secure online Bookshop.