Language, Science and Reference


…Suppose Peter says that Joe Sixpack voted for a living wage because he’s worried about his childs health. Are we entitled to conclude that Peter believed the world to be constituted of such entities as Joe Sixpack, living wages, and health, and relations like voting-for and worrying-about that hold among them? Would the parallel inference be legitimate when Peter says that Tom visited Boston? If Peter says that the bank moved across the street after it was destroyed by fire, does he believe that among the things in the world there are some that can be destroyed but still be around, so that they can move? (Chomsky: New Horizons in The Study of Language and Mind)

As the above quote shows Chomsky argues that if we use a term in language it does not follow automatically that the term commits us ontologically. So if we take terms  like: flaw, the average man, unicorn, it does not follow that because these terms exist in our language that they refer to mind independent entities.  He correctly stresses that we should not read off our ontology from our ordinary way of speaking. In another context he mentions how our ordinary concepts reflect intricate and surprising constraints on how we can interpret them, he also believes that these constraints are best explained interms of innate concepts. These constraints will to a certain extent determine how we use our concepts to refer.  Chomsky labels these constraints an I-variant of Frege’s telescope, implying that it is through the lense of these constraints that we can refer to entities in the world. So, for example, if we take the word ‘London’, this word has various different properties some of which are contradictory:

We can regard London with or without regard to its population: from one point of view, it is the same city if its people desert it; from another, we can say that London came to have a harsher feel to it through the Thatcher years, a comment on how people act and live. Referring to London, we can be talking about a location or area, people who sometimes live there, the air above it (but not too high), buildings, institutions, etc., in various combinations  (as in London is so unhappy, ugly and polluted that it should be destroyed and rebuilt 100 miles away, still being the same city). Such terms as London  are used to talk about the actual world, but neither are or are believed to be things-in-the world with the properties of  the intricate modes of reference that a city name encapsulates. (NHLM, p. 37)

So terms such as ‘London’ are used to talk about the actual world but, according to Chomsky, people do not think that there are things in the world with the properties of the intricate modes of reference that a city name encapsulates.  Chomsky’s discussion of the vagaries of reference here will remind people of Quine’s similar discussion in chapter 4 of Word and Object:

Insofar as it is left unsettled how far down the spectrum toward yellow or up toward blue a thing can be and still count as green, ‘green’ is vague. Insofar as it is left unsettled where to withhold ‘muddy water’ in favour of ‘wet mud’, ‘water’ and ‘mud’ are vague. In so far as it is left unsettled how far from the summit of Mount Rainier one can be and still count as on Mount Rainier, ‘Mount Rainier’ is vague. (Quine: Word and Object, p. 126)

In the above section Quine is discussing the vagueness of terms and of how the extension of the terms are unclear. Here Quine is reminding us that while some sentences may appear to be purely referential one of the terms in the sentence does not actually refer at all and is in fact a purely relational term. So clearly Quine and Chomsky are making similar points about language. The points being that ordinary language is shot through with vagueness and ambiguity and that while some terms may appear to be referential when analysed it turns out that they are not referential at all.

Quine emphasises that the vagueness of term is a natural consequence of word learning. So, for example, he claims that the indefinite objects of a vague term are the terms which bear only a slight similarity to the objects for which we have been rewarded for verbally responding to in the past.  So if we think of a child learning a word by induction through observing societys usage of the term, the vague cases will be the cases for which induction is most inconclusive because of a lack of evidence.  Quine argues that the evidence will not be there to be gathered because of society’s members having themselves had to accept similarly fuzzy edges when they were learning.

Vague words can he claims be either singular or general.  He gives the following example of vague terms:

Insofar as it is left unsettled how far down the spectrum toward yellow or up toward blue a thing can be and still count as green, ‘green’ is vague. Insofar as it is left unsettled where to withhold ‘muddy water’ in favour of ‘wet mud’, ‘water’ and ‘mud’ are vague. In so far as it is left unsettled how far from the summit of Mount Rainier one can be and still count as on Mount Rainier, ‘Mount Rainier’ is vague. (Word and Object, p.126)

According to Quine a general term true of objects can be vague in two different ways. Firstly because of the several boundaries of the objects, and secondly because of the ambiguity as to whether to include or exclude marginal objects.

So, for example, if we take a term like ‘mountain’ it is vague as to how much terrain we can count as a mountain, where exactly the cut off point is? It is also vague as to what lesser entities we can even count as mountains at all.

Ambiguous terms differ from Vague terms. A Vague term is a term which we are unsure how to apply to marginal objects, whereas an ambigous term may variously true of certain objects, and false of them aswell. Quine, gives the example of the term ‘light’,  this, he argues, will be true of ‘dark feathers’ but false of them in another sense. In the first sense we will take ‘light’ as ‘light as a feather’, in the second we will take ‘light’ as ‘a beam of light’. Quine claims that a partial explanation of ambiguity can be achieved by considering the linguistic term homonyms.  His gloss on homonyms is as follows:

Lexicographers and grammarians have long permitted themselves to treat words otherwise than as linguistic forms, by declaring of a form that it functions sometimes as one word and sometimes as another. (ibid, p. 129)

This of course leaves us with another question such as when do we have two homonyms as opposed to one ambiguous word? According to Quine lexicographers suit their convenience when trying to answer such questions. So if we take the example of ‘bore’, in the sense of ‘she bore her child’, and in the different sense of ‘she is a bore’ the lexicogropher will treat ‘bore’ in these sentences as homonyms because of their different etymology, and different grammatical functions.

Quine claims that we deliberately create ambiguity when we name somebody. He uses the example of naming somebody ‘Paul’, this name is not a general term true of many people it is rather a singular term with wide ambiguity. The evidence for this claim is based on a grammatical fact about general terms. A general term like ‘Dog’, permits the following grammatical constructions and it is the possibility of such constructions that marks out a term as a general term: a dog, the dog, dogs etc. Obviously in the case of ‘Paul’ we are not permitted to use such constructions as ‘A Paul’, ‘The Paul’, ‘Pauls’ etc.  The example of ‘Paul’ distinguishes between general terms and ambiguous singular terms in terms of the grammar of english.

Now though Quine asks us to consider a general term; and he asks us how we can decide how much of the terms multiple applicability is due to ambiguity and how much is due to generality? To decide between the two alternatives Quine gives some concrete examples of terms with multiple applications such the following:

[1] That is a hard question.

[2] That is a hard table.

[3] The proposition ‘either it is going to rain or it is not going to rain’ is true.

[4] The proposition ‘it will rain tomorrow’ is true.

[5] Material objects exist.

[6] Numbers exist.

Some philosophers claim that the above terms (in italics) are different uses of the same ambiguous terms. Quine disagrees and claims rather that they are merely very general terms:

Why  not view ‘true’ as unambiguous but very general, and recognize the difference between true logical laws and true confessions as a difference merely between logical laws and confessions?

Ultimately Quine suggests that we treat terms as being ambiguous when they can be clearly true and false of one and the same thing from utterance to utterance.  However he reminds us that while this may be our best way of discovering what terms are ambiguous it certainly does not follow that terms which vary in truth value can all be explained in terms of ambiguity. He gives the example of the statement ‘The door is open’ which changes its truth value depending on the movement of the door but which is not an ambiguous statement.

A further point which Quine makes is that ambiguity can have implications for composite words in the sense that such words can be interpreted in the attributive sense and the syncategorematic sense. Quine explains this distinction in the following manner:

One way it enters is through indeterminacy between the truly attributive and the syncategorematic use of certain adjectives. Thus consider the rich little word ‘poor’  When it is ostensibly in attributive position it may either have truly attributive use, in which case it may either impute poverty or express pity, or it may be syncategorematic, suggesting ‘badly’. If in  ‘poor violinist’ we take the use of poor as truly attributive, then poor violinsts are poor (or perhaps pitiable) and they are violinists; if we take it in the syncategorematic way,  then poor violinsts need to be neither poor nor pitiable nor even, by descent standards, violinists. (ibid p.132)

Quine further clarifys his views on this topic by explaining that when it comes to ambiguity of syncategorematic compounds that the ambigious term is the compound not the adjective; because an adjective in a syncategorematic use is not used as a term. Quine offers a good example of this type of sentence: ‘I passed nobody on the street’ where nobody is clearly not being used as a term.

Like Quine Chomsky also discusses homonymy:

Consider the word ‘Bank’  (savings, river). We can say that:

(1) The bank burned down and then it moved across the street;

(2) The bank, which had raised the interest rate, was destroyed by fire; and

(3) The bank lowered the interest rate to keep from being blown up.

Referential dependence is preserved across the abstract concrete divide. Thus  (1) means that the building burned down and then the institution moved; similarly (2) and (3). But we cannot say that:

(4) The bank burned down and then it eroded; or

(5) The bank, which had raised the interest rate, was eroding fast; or

(6) The bank raised the interest rate without eroding.

Sentence (4) does not mean that the savings bank burned down and then the river bank eroded.

The facts are often clear, but not trivial. Thus, referentially dependent elements, even the most narrowly constrained, observe some distinctions but ignore others, pronouns, relatives, the ‘empty category’ that is the subject of ‘being blown up’ and eroding’. In the case of ‘bank’, the natural conclusion is that there are two Lis that happen to share the same I-sound (homonymy),  and that one of them,’savings bank’ is polysemous, like ‘book’: it provides a way of looking at the world that combines abstract and concrete properties, allowing referential dependence across these perspectives. (Chomsky, NHLM p. 180)

Chomsky is not here concerned with how to distinguish between an ambiguous general term and homonyms, his concern is with showing the complexity of our referential apparatus and how subject it is to internal constraints.

So he asks us to think of the linguistic form ‘Bank’ this written inscription (or spoken sound) is used to label two different concepts: (1) Bank as in river bank, (2) Bank as in financial institution. The key point to note in the above 6 examples is the referential dependence of the pronoun it on Bank and how this dependence changes depending on how we interpret the term ‘bank’. Thus in consider 1 in relation to 4:

(1) The bank burned down and then it moved across the street.

(4) The bank burned down and then it eroded.

A number of points should be obvious from the above sentences; firstly in 1 ‘Bank’ and ‘it’ are referentially dependent on each other. ‘Bank’ in 1 refers to a concrete object i.e a particular building, whereas ‘it’  refers to an abstract entity i.e. a financial institute, nonetheless despite this fact ‘Bank’ and ‘it’ are referentially dependent on each other in sentence 1. Now let us consider sentence (4)  we cannot treat ‘bank’ as in financial institution in the first part of the sentence and treat ‘it’ as refering to river bank in the second part. It is a fact about referential dependence that such a move is blocked.

While both thinkers agree about the ambiguity of language they use this point for different purposes. Quine who thinks that our scientific theories are a web of sentences which face the tribunal of experience as a corporate whole tries to tidy up our language by translating it into the syntax of quantification. Chomsky, on the other hand thinks that a different faculty of the mind (our Science Forming Faculty) is responsible for us developing our scientific theories so he argues that we don’t need to modify our language. Our ordinary language is vague for Chomsky but it is ok for the purposes of day to day communication. Chomsky only discusses the problems of conceiving of language as referential to illustrate what he conceives of as a poverty of stimulus for concept acquisition.

Chomsky’s primary point is that the words of our language are richly constrained by rules of the lexicon and that the meaning of a term is not to be discovered by pairing it with an external object in the world. He is making the point that no external object could have the properties which we attribute to our concepts such as being concrete and abstract at the same time. He cites approvingly Strawsons clsim that there is no such thing as a logically proper name. And he alludes to the fact that any act of naming will require a lot of stage setting (he cites Goodman explicitely) and he claims that this stage setting will be provided by our innate constitution.

Chomsky asks us what goes into refering to an object, he claims that such things, as design, intended and characteristic use and institutional role will be a decisive factor in deciding. He claims that if something looks like a book, but we are told that it was designed to be a paperweight we may come to think of it as a paperweight.  We can think of a book as being an abstract entity as in ‘War and Peace’ or a concrete entity as in a particular copy of ‘War and Peace’.

Chomsky believes that the proceeding discussion shows that the act of refering requires intensive mental architeciture. Chomsky makes the following point about reference:

The question, ‘to what does word X refer?’ has no clear sense, whether posed for Peter, or (more mysteriously) for some ‘common language’. In general, a word, even of the simplest kind, does not pick out an entity of the world, or of our ‘belief space’- which is not to deny, of course, that there are banks… The observations extend to the simplest referential and referentially dependent elements [pro-nouns, same, re build, etc]; or to proper names, which have rich semantic-conceptual properties derived in large part from our nature, with some overlay of experience. (NHLM p 181)

In the above section Chomsky is claiming something which seems somewhat bizzare; he is claiming that a question such as ‘to what does word x refer?’ is senseless, yet at the same time he is claiming that there are of course things like banks in the world.  This passage is difficult to interpret because Chomsky seems to be claiming that a question, ‘To what does the word ‘bank’ refer to? Is a senseless question, yet on the other hand he admits that banks do of course exist. So a sceptic could ask; Given that banks exist, why can’t we just claim that ‘banks’ refer to banks? The obvious reply to the above question is that if we take the word ‘bank’ on it own and say that it refers to bank, does this mean riverbank, lending bank, the abstract entity bank as a finantial institution, lending bank as in a particular building? This can only be decided by seeing how the word is used in sentences (its syntactic constraints), and by seeing the contexts the word is used in (pragmatic use of the word). So the reference of the world is determined by how the person uses the word, and the way  the person uses the word. In other words people refer not words.

Chomsky follows up his remarks on how we refer to bank by claiming that the act of naming is a kind of ‘world-making’, in Goodman’s sense of the word. However Chomsky believes that the world-making is made possible by our innate human nature not by experience. Chomsky notes at this point that his approach to semantic-interpretation has a traditional flavor to it and he likens it to the approach which was taken by seventeenth-century rationalist philosophers. For Chomsky according to rationalist philosophers claimed that we recieved the data of sense and interpreted it according to our innate cognoscitive powers. So for both Chomsky and the rationalists our interpretations of data was largely achieved as a result of our innate prowess.

Chomsky then goes on to quote Locke, Hume, Hobbes and Cudworth as exemplars of the view on semantic-interpretation which he favours.  This itself is a curious move considering that with the exception of Cudworth none of the above thinkers are infact rationalists. However given that we know from other writings that Chomsky knows that Hume and Locke were empiricists not rationalists we will not interpret him as making such a basic mistake. Chomsky’s quotes from Hobbes, Locke and Hume can be read as exemplifying the complexity of the ordinary concepts we use such as those of: tree, river, person etc. Furthermore Chomsky seems to be implying that concepts which such rich intricate structure cannot be learned from experience despite what the empiricist philosophers seem to think.

It is perhaps helpful at this point to illustrate some of the points which Hume, Locke, and Hobbes made about according to Chomsky:

A man will always be the same, whose actions and thoughts proceed all from the same beginning of motion, namely, that which was in his generation; and that will be the same river which flows from one and the same fountain, whether the same water, or other water, or something else than water, flow from thence (as in the classical case of the ship of Theseus); and that the same city, whose acts proceed continually from the same institution. ( NHLM, p. 183 quoting Hobbes)

This quote which Chomsky attributes to Hobbes is that the nature of identity whether of ships, or rivers, or cities can be traced to the point of their origin. So for example we cannot say that the river liffey and the river boyne are the same river because they originated from different sources. Chomsky then goes on to discuss how philosophers from Locke to Hume developed the concept of personal identity:

The inquiry into personal identity from Locke to Hume was concerned with organic unity, a broader notion. A tree or an animal ‘differs from a mass of matter,’ Locke noted by virtue of the ‘organization of parts, in one coherent body, partaking of one common life’ with ‘continued organization’ that comes from within, unlike artifacts. The identity of an oak resides in ‘a sympathy of parts’ contributing to ‘one common end’ of ‘support, nourismhment and propagation’ of the form, Shaftesbury added. Hume largely agreed, though taking ‘the identity, which we ascribe to the mind of men,’ and ‘the like kind… that we ascribe to vegetables and animal bodies,’ to be ‘only a fictitious one’ established by the imagination, not Shaftesbury’s ‘peculiar nature belonging to this form’ (NHLM, p.183)

So Chomsky goes from speaking sympathetically of Hobbes view that identity is to do with the source of the thing considered to move on to consider later views on the topic. He notes Locke’s view that the identity of a thing is concerned with its organic unity, and notes that Shaftesbury added to this account that a further criterion of calling something an entity is that it has a common end, e.g. propagation of the form. It is the Hume’s view that these identities we project onto things are just psychological constructs which particularly interests Chomsky.

Chomsky agrees with Hume on this point and thinks that when we look closely at the intricacy of concepts we will come to see that Hume is correct. He cites the work of Locke, to illustrate exactly how intricate and detailed concepts are. A person for Locke was a forensic term which had the following properties:

Person is a forensic term appropriating actions and their merit; and so belongs only to intelligent agents, capable of a law, and happiness and misery as well as accountability for actions and much else. (NHLM, p.183 Chomsky quoting Locke)

Chomsky then goes on to quote Locke’s views on the identity of a river which go beyond Hobbes views on the matter. Chomsky claims that such views on identity (which he considers a psychological matter not a metaphysical matter) barely skim the surface of what we know of concepts and that our understanding of these concepts can only be marginally be effected by experience. Once we realise the complex nature of these innate concepts our task in semantics becomes studying their nature, and how they interact with other faculties of the mind to yield our rich interpretation of the world. Our task is no longer to study the relation between concepts and the world.

One of the first points which come to mind upon reading Chomsky’s discussion of concepts is his reliance on complex philosophical theories of identity as a way of revealing incredible complexity of our concepts of person, tree, river etc. Chomsky views these complex concepts as being psychological ways of organising our experiences, and he claims that such philosophical exposition of concepts hardly skim the surface when compared with the knowledge that people implicitely have. According to Chomsky most of our conceptual knowledge is unconscious, this can be seen from the fact that ordinary language users can reliably distinguish between rivers, trees, and persons, as well as speak about such entities in a manner that has coherence and appropriatness to the situation, yet still be unable to give philosphical explications of such concepts.  So, for Chomsky, philosophers like Locke, Hobbes and Hume are doing their best to try and rationally reconstruct what it is that they implicitely know. So Chomsky can be fruitfully seen as a sort of inverse Hume. Hume believed that our ideas were faint copies of our impressions, wheras Chomsky believes that our ideas (concepts, lexical items) are innate rules which we unconsciously use to interpret the flux of experience.

Chomsky has often written about his picture semantics asthough it were diametrically opposed to Quine’s view on semantics. Here he criticises what he takes to be a modern consensus in the philosophy of language:

In general, there appears to be no reference  relation in human language and thought in the technical sense of Frege, Peirce, Tarski, Quine, and contemporary ‘externalist’ philosophy of language and mind. Referring is an action, and the internal symbols that are used to refer do not pick out mind-independent objects. On investegation, it turns out that what we understant to be a house, a river, a person, a tree, water, and so on, is not a physical construct of some kind. Rather, these are creations of what seventeenth century investigators called our ‘cognoscitive powers’, which provide us with a rich means to interpret and refer to the outside world from certain perspectives. (Chomsky Evolution of Human Language: p, 12)

We saw above that Quine actually agrees with Chomsky on the ambiguities of language and the vagueness of its reference. So Quine doesn’t hold the unrealistic conception of language reference that Chomsky attributes to him. Where the real difference lies between both thinkers is that Quine doesn’t use facts about the ambiguous vague nature of language to argue for innate concepts.

Quine’s magum opus Word and Object is primarily concerned with two things (1) How a child develops his overall theory of the world through learning his language, (2) How we can regiment this language to make its ontological commitments clear.  When Quine speaks of language being a social art which we learn from in our peers in intersubjectively conspicuous circumstances, he seems to be endorsing precisely the view of language which Chomsky is criticising. But as always in philosophy, the situation is much more complicated than this.

When Chomsky speaks of  concepts of ‘tables’, ‘trees’, and ‘people’ as being psychological concepts which we use to organise our experiences he is echoing claims which Quine made 50 years eariler in From a Logical Point of View.  In his third paper of From a Logical Point of View, called Identity, Ostension, and Hypostasis Quine spoke about the process of imputing identity to an object. He emphasised how ostensive definition would be massively undetermined by the data:

Pointing is of itself ambiguous as to the temporal spread of the indicated object. Even  given that the indicated object is to be a process with considerable temporal spread, and hence a summation of momentary objects, still pointing does not tell us which summation of momentary objects is intended,  beyond the fact that the momentary object at hand is to be in the desired summation. Pointing to a, if construed as referring to a time-extended process and not merely to the momentary object a, could be interpreted either as referring to the river Cayster of which a and b are stages, or as referring to the water of which a and c are stages, or as referring to any one of an unlimited numner of further less natural summations to which a also belongs. (LPV, p.67)

In other words ostensive definition requires a lot of stage setting. Quine points out that such difficulties are typically overcome by saying ‘this river’ in other words appealing a prior concept of river as one distinctive time consuming process, one distinctive form of summation of momentary objects.  At this point Quine notes we have moved beyond ostension and moved to the stage of conceptualisation. However he then asks us to assume that we do not already have the prior concept of ‘river’, and asks again how far we could get with ostension alone?

Quine sketches the following procedure:

What we may do then is point to a and two days later to b and say each time, ‘This is the Cayster’.The word ‘this’ so used must have referred not to a nor to b, but beyond to something more inclusive, identical in the two cases. Our specification of the Cayster is not yet unique, however, for we might still mean any of a vast variety of other collections of momentary objects, related in other modes than that of river kinship; all we know is that a and b, are among its constituents. By pointing to more and more stages additional to a and b, however, we eliminate more and more alternatives, until our listener, aided by his own tendency to favor the most natural  groupings has grasped the idea of Cayster.

Quine’s emphasis on ‘natural groupings’ shows that he allows some innate apparatus to explain concept learning, however his primary emphasis is reinforcement and trial and error.  Chomsky  on the other hand when discussing the complex ordinary language concept of a house simply uses the complexity of the concept to show that the concept must be innate.

Chomsky’s discussion aims to clarify what the meaning of the word is, and to emphasise how much information is included in such a supposedly simple concept,  he thinks that this complexity shows that the concepts cannot have been learned in a purely referentialist manner. One interesting property of the concept ‘house’ is that if we were to say that ‘John is painting the house’, we mean that he is painting the exterior of the house not the interior. Another connected feature of the concept of ‘house’ is the concept ‘brown house’, Chomsky notes that a universal feature of language is that ‘brown house’ will be interpretated as meaning that the exterior of the house is brown not the interior. One point to note about this claims Chomsky is that exterior of a surface is very important for understanding a term. Thus if I see a house I see its exterior surface, seeing its interior surface does not surfice. But he notes a house is not just its exterior surface, a geometric substance. If John and Mary are equi-distant to the exterior of a house but John is in the house and Mary is outside the house. We will say that Mary is near the house but not that John is. But he  notes the interior of  the house is abstractly concieved, the house will be the same house even if I fill it entirely with concrete.

These strange properties complex properties of the complex concept of house combined with the speed children supposedly acquire the concepts are used as evidence that the concepts are innate. I will not go into who is right and who is wrong on the question of concept acquisition (spoiler alert: neither of them are), here I merely wanted to show that both thinkers agree that our ordinary language is too vague and sloppy to be accorded any ontological significance. This is a problem for Quine who thinks that our overall scientific theories of the world are an interconnected set of sentences. Hence Quine has to try and modify language by translating it into the syntax of quantification. Chomsky who thinks our science forming faculty has little to do with our language faculty.

In my next blog I will discuss why Chomsky’s postulation of a science forming faculty has little evidence to support it. In the following one I will show that Quine’s attempt to translate scientific theories into the syntax of quantification has failed. Given that both of their explanations of scientific competence fail and their views on language show that language is too vague to house the precise theories of science, neither has a satisfactory theory of scientific competence. So in my final blog I will discuss how Paul Churchland’s (as interpreted by Matt Bush) theory of Hebbian Maps solves some of the difficulties facing Chomsky and Quine.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s