In the matter of reviews - Adam Sennet

[2] These comments follow those on Koslicki's review. Sennet’s review may be helpful to the reader of Words without Objects, in that it reviews the book on a chapter-by-chapter basis - albeit rather briefly, given the constraints of the exercise. It is also in some respects more contentious than Koslicki’s: Sennet mentions a number of considerations which could perhaps turn out to count against the thesis of Words without Objects. At the same time, given their obvious brevity and informality, it seems fair to assume that these are less intended as actual than as possible or potential objections, waiting to be suitably developed and expanded.

For instance, Sennet seems to think — and I stress the ‘seems’ — that what I argue to be the semantical non-singularity of a definite non-count description can be summarily refuted. Thus, so he maintains, given that the description ‘my favourite thing in the world right now’ is construed, not as the sort of idiom which we would naturally take it to be, but in all semantic seriousness, as designating a single unit of some kind, and that the subsequent ‘is’ in the sentence

My favourite thing in the world right now is the whiskey in my glass

is one of identity, then the invocation of Leibniz’ law is sufficient to establish the semantical singularity of the further description ‘the whiskey in my glass’. The idiomatic aspects of English syntax do not necessarily embody this sort of semantical seriousness, of course; ‘My favourite topic in astronomy is the moons of Jupiter’ can surely be allowed to pass for true without committing the speaker to a set-theoretical construal of the plural.

However, the argument would seem to beg the question. Thus Sennet takes it for granted that (on his semantically serious construal of the initial description) the sentence in question could, at least in principle, be true. Yet for reasons with which Sennet is at least in broad outline familiar, that assumption is one which I am unable to accept. Given the existence of sound arguments for the assertion that definite non-count descriptions are semantically non-singular (such as I certainly claim to present) no identity statement involving a definite semantically non-count description and a semantically singular description could possibly be true.

A related proposal which Sennet appears to take seriously is the following: even semantically genuine singular descriptions can fail to meet the demands of Russell’s theory of singular descriptions (RTD). Thus suppose a red patch on my wall, such that ‘the red patch on my wall’ successfully denotes . Yet any proper part of this patch is itself a red patch — ‘presumably’, as Sennet rather cautiously remarks — hence the Theory of Descriptions must be just plain false, Q.E.D. But to simply ‘presume’ that (given a non-stipulative, natural-language meaning for 'patch') there is not exactly one red patch, when there is exactly one non-discontinuous red-coloured region of space on the wall, seems to me a remarkably insubstantial basis for the outright abandonment of RTD.

What however I would especially like to comment on is a remark of Sennet’s which I suspect to be symptomatic or representative of a more widely held view of non-count nouns. Thus, in concluding his reflections on RTD (and the defense of RTD in Words without Objects), Sennet opines that ‘atomicity’ — and in particular, the lack of it — ‘is really at the heart of the matter’. Or in other words, 'the key difference' is that between NCNs such as ‘furniture’, which although non-count, are nevertheless in virtue of their semantics true of a range of what I have called atomic units (in this case, pieces of furniture), and other NCNs such as ‘wine’, which are true of no such units, or are linked to no cognate individuating expressions. Here, then, it seems to be implied, the contrast of CNs and NCNs shows its ‘true colours’, or in other words, its ‘ontic stripes’. Only here, perhaps, do NCNs show their true indispensability or irreducibility.

But now in the first place, the thought is that there is a ‘heart of the matter’ — exactly one heart, that is — a certain underlying core or key to the entire ‘mystery of mass nouns’. And secondly, there is the thought that this core is ontological. On such a view, ‘the problem’ reduces to a single issue, an issue of ontology. The idea is an attractive one, and it has attracted a substantial body of support.

For several related reasons, however, this strikes me as a dangerous mistake. It is a central thesis of the book that there is no such single issue, no such heart or core — and that, instead, there are several quite distinct, related but equally significant ‘keys’ to unravelling the issues surrounding NCNs. (I wholeheartedly endorse the spirit of Aristotle’s remark, that ‘being is said in many ways’ - look at the network as a whole, and at the totality of relationships between its parts). The fact that NCNs as such do not specify countables, even though this is a purely semantic, and not an ontic fact, is I believe no less important to understanding this group of philosophical issues, than is the fact that there are indeed ontically distinctive groups of NCNs. Furthermore, to cast the issue as one of atomism inclines one to overlook that range of NCNs which both are and are not ontically distinctive — that is, NCNs which do carry an implication of discrete units or particulate composition, but which carry no implication of atomicity. ‘Sand’, ‘gravel’, ‘snow’ and ‘grit’ all have cognate CNs — ‘grains of sand’, etc. — but carry no implication of atomicity; grains can be subdivided (and for Sorites-like reasons there is no smallest grain). Hence in contrast with ‘furniture’, ‘footwear’, and their kin, but in harmony with ‘water’, ‘wine’ and so on, there is here no question of a priori grounding issues of identity on constituent particles.

When all is said and done, Sennet’s review comes across as essentially tentative. But given the unfamiliar character of the thesis of Words without Objects , this is probably a reasonable strategy. As to whether that thesis is fundamentally sound or not, Sennet, perhaps wisely (if not altogether comfortably) opts to sit squarely on the fence.

the matter of reviews - Kathrin Koslicki

[1] It may appear a tad narcissistic to begin a whole new blog with some remarks on reviews of one's own recent book. But by way of mitigation, I see this medium as an opportunity for self-expression, akin to a diary, but subject to the objective constraint of potential public scrutiny.

Be that as it may, the fact is that the bloggosphere seems the ideal place for minor comment or 'intervention' concerning the state of play on issues of some academic interest. And by 'minor' comment I mean, among other things, non-refereed publication.

And so, without further ado, I will launch into just such remarks. These remarks address two extant reviews of Words without Objects, one by Adam Sennet, the other by Kathrin Koslicki. Both reviews do me the honour of taking the book seriously -- really engaging with its unfamiliar train of thought. I must express my gratitude to both Adam and Kathrin on this account.

(Koslicki's review is in the Australasian Journal for March of 07, and Sennet's is in the Notre Dame Review of Books, also for March of 07. Both are available online).

These two reviews are complementary, in that while Sennet's tends to focus on the framework and logical dimension of the book, Koslicki's addresses mostly its semantic and metaphysical aspects. I comment on Koslicki's review first.

1. Koslicki is exactly right to say that the primary semantic aim of the book is to 'give an analysis of the phenomenon of non-singularity'; and I'd like to take the opportunity here to thank Alan Sidelle in this connection, for his invaluable help in focussing me clearly on that aim.

But Koslicki is especially interested in what I like to call pure non-count nouns ('pure NCNs') which are also, perhaps, the central focus of interest for those who speak of 'mass nouns' -- words like 'air', 'water', 'mud' and 'gold'. And since this is not only a relatively obscure region of our discipline, but also a 'mere blog', which precious few folks are likely to pay much attention to, I may as well point out that one unfortunate consequence of 'mass noun' talk is that it tends to obscure one of the most fundamental aspects of these nouns -- the glaringly elementary fact that they are not count nouns. What this very humdrum observation ought to hint at or suggest, it seems to me, is the remote possibility that what they apply to is not countable. (In fact, something rather similar to this crops up in the nondenumerability of the putative points in a volume of space or for that matter of the real numbers).

However to continue. Koslicki puts me -- not perhaps unfairly, it must be said -- in the prima facie embarrassing and almost pathological or megalomaniacal position of maintaining that since the demise of the Presocratics, we philosophers have all suffered from 'something like a collective delusion'. To borrow a term from Tom McKay, the delusion in question is Singularism. That's the compelling tendency to imagine, once again as philosophers, or more generally as reflective thinkers, that whatever we talk about must at least be 'singular' or one, that is, it has to be a unit and in that minimal sense an object. ('Singularism' is a tad unfortunate, in that the doctrine goes far beyond linguistic or semantical matters; but let that pass).

Well, whether that is true or not, there's no doubt in my mind that Meinongianism in one guise or another -- the tendency to 'posit objects', things quite unheard of outside the Ivory Tower -- is a real and enduring hazard in philosophical thought -- hence the tremendous importance of Russell's (and to a much lesser extent Quine's subsequent) attack on it, and the immeasurable if limited value of the 'logically transparent language' ideal. This Meinongian impulse may well have been correctly diagnosed by Wittgenstein - the reflective but nevertheless unwitting application of a 'basic' or simplest case of some phenomenon (here, the grammatical subject-expression, the simplest semantic form of which is precisely reference to some object) as model or paradigm for interpreting that phenomenon in all its other forms (here, non-singular subject expressions, quantified subject expressions, abstract subject-expressions, etc. etc. etc.). (But again, to balance these considerations on philosophical reflection, considerations of a practical nature which give centrality to individual object-thinking might also have some role to play. As Michael Ayers has noted, 'the apprehension of material structure, and of such discrete and unitary material structures, is enormously important for the simplest of actions. It is a condition of literally finding a way through the world, not to speak of manipulating it.... Ordinary objects, ordinary language, and identity)

It's for reasons of this sort that I would after all wish to stress semantics, among other things, over metaphysics. Koslicki is inclined to take my central claims in WWO to relate to the metaphysics of pure NCNs. And as she reasonably says, she 'fails to see how there is a sufficiently substantive difference here between the so-called 'pure' NCNs like 'air', 'water', 'ice', and 'mud', and the 'non-pure' NCNs, like 'snow', 'sand', 'furniture', and 'clothing', to warrant the postulation of an entirely distinct, non-objectual and non-identity-involving, semantics, logic, and metaphysics'.

But this seems to me to involve a not-insignificant misunderstanding. For the fundamental threat to the object-ontology is not metaphysical, and does not enter in with pure NCNs -- or even, for that matter, with NCNs of any sort whatsoever. The fundamental threat to the object-ontology is one which arises within the very framework of the object-ontology itself, in the phenomenon of non-singular reference to objects (reference in the plural, that is). For here, we have a form of reference which is not - if Boolos and others (including I myself) along with just plain old common sense - are right, reference to an object. (The 'delusion' that it is -- reference to some sort of collective entity or 'set' -- seems to be a purely philosophical or reflective delusion. 'The many' are not one. Philosophers regularly get things wildly wrong, in ways that plain folk would be inclined, in my view, to find quite astonishing).

And it is precisely here, paradoxically enough, that a 'non-objectual' semantics gets off the ground. If a non-singular reference, whatever its stripe, is not reference to an object. then it looks as if the difference between non-singular reference which is plural and non-singular reference which is not plural is precisely the difference between something which is reference to objects and something which is not - qua non-count reference - reference to objects! (This does not mean that there can be no objects in the extension of the expression; whether there are or not is not a function of the non-count status of the reference itself).

This might just strike one as somewhat naive -- and there's a sense in which it certainly appears innocent -- but it is definitely not rocket-science. I'll return to this topic in the next little while.

2. Koslicki speaks, cautiously, of 'the (alleged) "thing" / "stuff" dichotomy', as well as of 'the apparently ontological category of "stuff"'. Now the issue here is close to being the other side of the coin of section 1 above: there is quite certainly no ontic dichotomy of 'things and stuff', nor -- unless one artificially or stipulatively restricts the natural-language application of the expression 'stuff' -- is there an ontic category of stuff. 'Stuff' is perhaps the most general noun in the English language; it applies for instance to the stuff in my office, which is mostly books, papers, pencils, pens, items of clothing, cups, mugs, photographs and the like - and it also applies to the stuff on the floor of my office, which at the moment is, among other things, dried up coffee, small amounts of post-winter mud, grit, dust and so forth, as well as a large amount of styrofoam packing-material still sitting in a box. In other words, the term -- a highly idiomatic one, to be sure -- is an equal-opportunity expression both for things of all sorts, collectively, and for materials or substances of all sorts as well. (The Aristotelian tradition has really ruined the unapologetic use of the very nice word 'substance'; but what can we do?)

What the 'the (alleged) "thing" / "stuff" dichotomy' really does represent, I'd suggest, is a contrast which does have a semantic correlate -- the humdrum contrast of measuring and counting. Describing the things in my office as stuff I do not need to concern myself with how many items there are, but only, for example, how much space they collectively take up, how much they collectively weigh, and so on; I treat them in effect as on a par with the styrofoam, grit and mud.

It seems to me important to recognise that 'stuff', as the tremendously general and non-ontic expression that it is, nevertheless leaves open a space within its scope for more plausibly ontic sub-classes of expressions. (And to impose the object-concept on such classes -- to talk in terms of 'parcels', 'quantities' or 'masses' -- is tantamount to an idealism of sorts, albeit idealism in its most materialistic of guises). But the die for the semantics of these ontic sub-classes is cast at the very general non-ontic level -- a level the semantics of which is itself 'non-objectual', in calling not for concepts associated with the discrete natural-number counting system, but for concepts theorizable via the continuous / dense real number system.

End of this particular chunk; I will comment on Sennet's piece in the next post.