Innate Ideas

Nativism in the Light of Locke’s Critique on Innate Principles

Wolfgang Unger

Term Paper in Phil 702, Locke’s Essay

Instructor: Prof. Vere Chappell

Department of Philosophy, University of Massachusetts, Amherst

ABSTRACT

The content of this paper is a critical discussion of the ideas of modern nativists and their relation to Locke's criticism of innatism. I investigate whether the arguments against innatism proposed by Locke are still effective against nativism.

This paper is divided into two parts. In the first part, I introduce the ideas behind nativism and put them into their historical context of the thoughts of continental rationalism. In the second part, I turn to specific theories within nativism and face them with Locke's arguments. In short, I argue that Chomsky can be interpreted as stating that not beliefs about grammaticality, but the principles underlying these beliefs are innate. Moreover, these principles are represented in the human mind as faculties rather then dispositions.

PART ONE

The Two Aspects of Empiricism and Rationalism

If one wishes to contrast nativism with empiricism, it is necessary to distinguish two different and independent issues of empiricism: First, there is the epistemological issue, concerned with the thesis that beliefs are justified through experience, i.e. through relying on sense data. In a radical version, this not only applies to beliefs about the external world, but also beliefs about abstract entities, since such beliefs, according to some empiricists, are just generalizations of concrete experiences.[1] The opposing theory to epistemological empiricism is rationalism - more specific: continental rationalism, which claims that reason has a role prior to experience, and that many principles are knowable (i.e. justifiable) by introspection, without making use of experience.

Second, there is the genetic issue, concerned with the origin of our ideas: Where do our beliefs and ideas come from? Whereas the empiricist’s answer of course is that we obtain all our ideas from experience, the rationalist holds that at least some of our ideas are innate. This view is called innatism. John Locke addresses in his “Essay Concerning Human Understanding”[2] mainly this issue when he is attacking the innatist’s account of ideas. One might object that there is no need to tie together rationalism and innatism, since rationalism does not necessarily imply innatism. But as a matter of fact, in the history of philosophy all important rationalists (the continental rationalists are meant) held both views. And this is certainly no coincidence.

In his paper “Innate Knowledge”, Alvin Goldman investigates if being innate has any justificatory quality. He comes to the conclusion (within the frame of an externalist account of justification in terms of a causal connection between the fact that p and the belief that p) that innate principles can indeed be justified just in virtue of being innate: he argues that evolutionary adaptation is a causal connection since being born with a certain belief can be helpful to survive[3] (and the evolutionary process, in his view, would of course favor the true beliefs). But such an account of justification is highly controversial, and most epistemologists instead hold the view that being innate makes beliefs not justified.[4] We can easily imagine that innate beliefs are false.

The rationalists hoped to show that the origin of our innate principles somehow guarantees their truth (e.g. in Descartes, God as the origin of innate ideas is perfectly trustworthy), but this attempt was mistaken: In fact there are two problems with innateness as justification: (1a) how can we know that an innate belief is really true? And (2a) how can we know that a belief is innate, i.e. what kind of intuition accompanies innate beliefs? Both questions are not answered in Goldman’s proposal. Even if a highly reliable mechanism (such as God’s imprinting principles) existed, (1b) we were still not justified in concluding the truth as long as we are not justified in believing that there is such a mechanism, and, applied to a specific believe, (2b) we could never be sure that a belief is innate since our intuition is a bad measure of credibility. The latter was also objected by Locke, stating that so-called innate ideas are not clearer perceived than others.[5]

So I will put aside the epistemological issue completely and adapt Stich’s view[6] that to characterize a piece of knowledge as a priori (as it was often done by the rationalists[7]) is to say something about its justification, and to characterize it as innate is to say something about its origin. The term “innate knowledge” I will therefore often replace by “innate belief” to emphasize that justification is not automatically granted. But we should always keep in mind that “innate principles”, were one of the cornerstones of the philosophy of continental rationalism, and it is doubtful if innatism could be separated from rationalism at all.

Nativism as a Modern Version of Innatism

Nativism is a modern view rooted in innatism. The advocates of nativism are mainly philosophers who also work in the field of cognitive psychology or psycholinguistics: most notably Noam Chomsky and Jerry Fodor (although the latter adopted a more critical attitude towards nativism in his latest writings). I will focus in this paper on Chomsky’s works, assuming that his thoughts are most representative for nativism. The nativist’s general objection against empiricism is still the same as was raised by the rationalists: the human mind of a newborn child is no tabula rasa at all, but equipped with an internal structure.

But there are indeed great differences between 17^th century innatism and modern nativism, mainly grounded on the progress in biological and psychological research: whereas innatists assumed that God placed innate ideas and principles in the human mind, nativism is grounded on genetics. The nativists hold that innate beliefs are in some way genetically programmed to arise in our mind, that is to say that innate beliefs are the phenotypes of certain genotypes that all men have in common. Another point is that nativists regard psychological processes as computational processes, i.e. rule governed manipulations of metal representations.[8] Therefore, accounts for learning are nowadays given in terms of networks and distributed processing, a framework the continental rationalists could not even dream of.

There are two advantages of nativism in contrast to innatism: First, nativists claim to explain why certain beliefs are innate and others are not, whereas innatists can only refer to Gods unlimited wisdom, claiming that he knows best with which innate ideas and principles a human mind should be furnished. Moral principles are therefore ruled out by genetics (since social norms are rather determined by cultural factors than by genes), and do not play a great role in nativism. The innatist, however, can not provide a causal explanation as the nativist does. This – at least in principle – allows the nativist to verify if a certain belief that is proposed to be innate really could have arisen in the process of natural selection.

Second, nativists do not assume that innate beliefs are literally imprinted in the mind from birth. Instead they hold that such beliefs arise during the development, because of mechanisms that favor these beliefs. Their distinctiveness and therewith innateness is then rather a matter of degree and not a matter of make or break. This feature in nativism cuts off Locke’s argument that “imprinting any thing on the Mind without the Mind’s perceiving it, seems […] hardly intelligible.”[9] Nativists are not committed to the view that innate principles must be perceivable. Locke argues that notions which are not naturally imprinted could not be called innate. What Nativists have to show is then in what sense their innate principles/faculties/capacities still deserve to be called innate. Certainly the account of innateness will differ greatly from the account given by rationalists.

To some extent, nativism might be considered as uncontroversial, since nobody will doubt that the human mind is a product of both our genes and our experience, and that certain beliefs (e.g. the belief that a smiling face is an indication for happiness) automatically arise at some stage in the human development. On the other hand, it is far from obvious what factors affect our development, and we should certainly reject the view that our development is determined by our genes alone. The phenotype does not entail any complex behavior. Therefore, we should consider both, genes and experience, nature and nurture, as interacting causes which produce the beliefs and ideas of men. The crucial question then is which of these factors are more relevant in specific types of learning (such as language acquisition). According to Fiona Cowie, nativists then hold that innate factors play a more important role than environmental factors. But I doubt that this question can be answered in general. Instead, one has to investigate certain acquisition processes, determining the relevant factors.[10] This is what Chomsky attempted for language acquisition.

Innate Beliefs as dispositional Beliefs

As is indicated by the developmental character of innate beliefs, nativists regard innate beliefs as dispositional, or even predispositional, i.e. dispositional in advance. Therewith, the nativist does not claim that innate beliefs are at any time propositional. If a person does not reflect on his innate beliefs, he will not hold these beliefs explicitly. This, however, is an old thought, already discussed in Plato's Meno, where it is claimed that innate principles are often forgotten, but can be recollected through introspection. Also Descartes emphasized the dispositional character of innate ideas in comparing them with hereditary diseases. Stich points out[11] that Descartes' analogy interprets innate ideas only as a counterfactual condition from birth, and that it is necessary to determine the normal development under which the counterfactual condition turns into a dispositional condition. But here, so Stich, we are confronted with the difficulty to explain what a normal development is like. We do not need to go into the details, but it is plain that Stich's objection also applies to nativism.

There is one aspect in which setting aside justification for innate beliefs is advantageous: one is not committed to the view that the human mind ever has or will reflect on innate beliefs (which is a necessary condition for many internalist accounts of justification, like access internalism). Therefore, such a belief can be hold without knowing that one does hold this belief.

But the whole view that innate ideas are dispositional makes nativism much weaker and less exciting, even vacuous, as Stich writes.[12] What is interesting about beliefs which we have only the disposition to hold? What distinguishes innate beliefs from other beliefs that most men probably will hold at a certain time of their life (e.g. the belief that one is 10 years old)? Furthermore, the talk about dispositional beliefs is also not very compelling: as Stich points out, beliefs always require concepts, and having concepts entails having beliefs about the concepts. But this requirement can not hold for dispositional beliefs since one cannot have the disposition to form a belief if the involved concepts are not known yet. Rather, one either has an actual belief or none at all. As soon as I discuss more concrete positions I will contrast both views in greater detail.

Nelson Goodman even argues that viewing innate beliefs as dispositional must be wrong and cannot give an account of language acquisition:[13] It is absurd to say that a stone having the disposition to fall does learn to fall due to gravity. Similarly, answering in a fixed way to certain stimuli is not an appropriate view of the mind, so Goodman. Furthermore, concerning ideas, if innate ideas were not necessarily accessible, they would be just dispositions and no ideas at all.[14] Therefore, talking about dispositions is highly misleading.

Innate Beliefs as Faculties

Another proposal also made by the innatists is to consider innate beliefs as corresponding faculties. This of course is a very abstract notion of beliefs. The view suggests that faculties generate beliefs, and in this sense the beliefs are predispositional. But as was pointed out by Harman, there is a danger to confuse knowing how with knowing that, a problem I will discuss later concerning Chomsky. Instead of faculties, one can also talk about the principles underlying the mechanisms of such faculties. This picture explains in what way faculties can generate beliefs: these beliefs are just derived from the unconscious principles that are employed in a certain faculty. But this account certainly will not work for any faculty: If one knows how to play pool he will not necessarily also know the laws of mechanics which describe the process relevant in playing pool. Therefore, there has to exist a particular relation between innate faculties and the generated beliefs. (Playing pool or riding bicycle does not generate any beliefs at all.) The problem we have to deal with is not the claim that certain faculties are inborn – a claim every empiricist will certainly agree to, since e.g. perception is just such a faculty. But the empiricist will deny that such faculties can be equated with principles. Locke found that talking about innate principles as capacities is already a confession that there are no innate principles.

Empiricist's and Nativist's Accounts of Learning

In the literature there is often expressed the worry that the whole debate between innatists and empiricists is just a metaphorical talk and that it is difficult to see in what they are disagreeing about.[15] (Robert Adams for example characterizes the only difference between Descartes and Locke concerning their theories of origin of ideas in ascribing metaphysical economy to Locke and metaphysical generosity to Descartes.[16]) The objection is not only proposed against the rationalist’s metaphors (Descartes’ light of reason, Leibniz’s block of marble whose veins guide the sculptor in shaping), but also against the nativists: as we will see, Chomsky makes use of a computer-metaphor, characterizing the human process of learning as an input-output process. To challenge these worries, we will turn to the very important issue of how beliefs are acquired, since empiricists and nativists provide different answers that allow us to distinguish their views better.

The empiricist holds that all learning is performed by principles of association. Associative models employ learning mechanisms such as induction, association or conditioning. Instead of innate ideas they employ “combinatorial devices, putting together items from experience.”[17] The main factor which causally determines the effectiveness in learning is the frequency of presentation of such items. Such associative models are classically described with associative networks: each concept is presented by a node with a number of fans that connect this node with other nodes, representing association. Learning consists then in setting up new connections or resetting the weights (of the connections) and thresholds (of the nodes) which determine when a node is sufficiently activated to fire. This of course is not a neurological picture of what happens in the brain (since it is absurd to think that each concept is represented by a neuron) but an abstract mental model. In contrast, nativists hold that sense experience activates dispositions and transforms latent unperceived ideas into actual clearly perceived ideas.[18] To account for this, they favor instead connectionist’ models, i.e. models of distributed presentation: concepts are represented by a pattern of activation involving lots of nodes. But it is far from obvious why nativists favor these models. In fact, connectionism is a very broad notion in cognitive psychology, and even associative models can be phrased in terms of connectionism. Many psychologies nowadays favor connectionism. Connectionist models suggest a new way in which ideas/principles can be innate.[19] Its advantages is that it rules (maybe innate principles) are implemented in such models in a natural way (as we will see, e.g. the recursive character of Chomsky’s generative grammar can be explained in terms of a connectionist algorithm). Moreover, connectionism is more realistic as a mental theory (providing natural realism) since it better describes how mental processes work in neurological terms, e.g. distributed processing[20], and therewith connectionism proposes a unified theory of mind and brain. But it has to be emphasized that in favoring connectionism there is not a choice implied between empiricism and nativism. Therefore, we have to contrast associative models of learning with nativist models both based on connectionism.

Getting to the biological rules of learning improves the understanding of mental models and has the advantage that the empiricist’s and nativist’s accounts of learning become very specific and allow predictions that can be tested. On the other hand, there is the danger that we grow apart from the philosophical issues. It is time return to them now.

PART TWO

Language acquisition

We now turn to the more specific question how language is acquired. This subject is the only one that is to some extent developed by nativists, and this for two reasons: First, language is closely related to the notion of innateness (since, after all, innate beliefs are – as soon as they are made explicit – propositional). Only in language learning the two aspects “knowing how” and “knowing why” are relevant. Second, it seems that psycholinguistics provides us with a couple of impressive facts about language acquisition. From these facts there are some arguments derivable in favor of nativism. There are two interesting issues about language acquisition here: first, Chomsky’s hypothesis of universal grammar (which is an old idea – e.g. Humboldt’s “grammaire generale” – but was convincingly proposed by Chomsky, based on his theory of generative grammar). Second, Fodor’s view about concept acquisition. Whereas Chomsky deals mainly with the issue of innate principles, Fodor deals with the issue of innate ideas.[21] In the following I will focus on Chomsky, since he proposes the more interesting hypothesis.

The arguments in favor for nativism about language acquisition

According to Fiona Cowie, there are two distinct families of arguments leading to two independent versions of nativism.[22] First, the arguments from poverty of the stimulus are stating that the content of our mind is far too rich to come from the external world. Second, the arguments from impossibility are stating that the human mind is by far too complicated and mysterious as that empiricism could give an account of how ideas are acquired and therefore certain mental items must be inborn. Chomsky argues more along the line of the arguments from poverty of the stimulus, whereas Fodor argues along the arguments from impossibility.

Cowie identifies the arguments from poverty of the stimulus with a version of task-specific nativism: inborn intellectual skills are domain specific (whereas the empiricist holds that these skills are of general purpose, as it is the case in making use of induction). This leads to the thesis of “cognitive architecture”, attributing several inborn mechanisms to certain mental faculties. Language acquisition is then one among other faculties which go beyond empiricist’s accounts of learning. And she identifies the arguments from impossibility with a version anti-naturalistic nativism: psychology cannot give an account of learning at all, since it is a causal, antirational process, far too complex to be analyzed in psychological terms, which is the reason why empiricism must fail. I do not think that both arguments are separable in the way Cowie suggests. In fact, also Chomsky employs the argument from impossibility in a certain way (without holding an anti-naturalistic view). However, Cowie’s suggestion hits an important point of Chomsky’s theory: language acquisition employs language-specific innate grammatical rules that are not used in other learning processes, and the learning mechanisms are therefore task specific.

Others have objected[23] that it is implausible to claim that language acquisition is more like imprinting than other cases of learning. Maybe the rules which govern language acquisition and generate the knowledge about grammar are the same than are used in other acquisitions as well.

There is a general problem with arguments from impossibility: they try to show that empiricism in impossible instead of giving a nativist’ account of how language acquisition is possible. But also Locke argued similar, but in favor for empiricism, showing that innatism is not possible. Therefore we have to be very cautious with such arguments.

Evidence for the Innate Hypothesis

According to Noam Chomsky, the contemporary research in psycholinguistics supports the innate hypothesis, i.e. that there are innate principles producing beliefs/ideas that humans have the disposition to believe.[24] These principles are in such a way that they allow explanations for certain phenomena in language acquisition which associative models of language acquisition fail to explain. The phenomena which are held to support the innate hypothesis are the following:

(i) The Creative use of language[25]: men are able to make up a seemingly infinite number of meaningful sentences. This variety of sentences comes along with a variety in grammatical structures, too complex to be learned from experience.

(ii) Language acquisition of small children is astonishingly fast. In particular, language skills need not to be reinforced very much: there are cases in which a child just listens to its parents and then suddenly starts talking almost fluently.

(iii) Language skills seem to be largely independent of a person’s intellectual performance. Almost everyone masters the very complex process of language acquisition.

(iv) A child can learn any human language as a first language if it is brought up among people speaking this language.

Chomsky’s main point is that we must attribute language-competence to speakers far beyond what could have been learned. He writes: “The idea that sentences or sentence-forms are learned by association or conditioning or training […] is entirely at variance with obvious fact.”[26] This conclusion entails what Hillary Putnam called the “what else?” argument.[27] He objects that Chomsky argues in favor of the innate hypothesis because he cannot imagine what else could account for language learning. He rules out empiricism as a theory of language learning. All the above evidence therefore only suggests that the contemporary theories about language acquisition are deficient, and hence does not support the innate hypothesis. Putnam gives series of other arguments showing that the evidence (i) – (iv) can also be explained without employing the innate hypothesis. For example, he remarks that there are other skills apart from language skills that also need no reinforcement. Chomsky replied to Putnam’s critique in his article “Linguistics and Philosophy”, as I think with success, so I will not discuss their controversy here. But I think that there is some legitimacy for Putnam’s conclusion: “invoking innateness only postpones the problem of learning: it does not solve it;” which again invokes the suspicion that talking about innateness is neither necessary nor helpful.

An Approach to Innate Principles: Universal Grammar

Chomsky defends the innate hypothesis in terms of an elaborated linguistic theory. This theory shall show how the two representations of the semantics of a sentence – surface structure and deep structure – are related with each other. His claim is that very heavy conditions on grammar are universal, i.e. can be found in the grammar of all human languages. We need to make a brief excursion into Chomsky’s linguistic theory to follow his arguments.

Chomsky (Aspects of the Theory of Syntax, 1965) developed so called generative grammars (in contrast to traditional grammars), which consists of rules making use of recursive structures (e.g. “rewriting rules”). Such a grammar has three components: a syntactic component (which consists of an infinite set of abstract formal objects), a phonological component (which relates syntactic structure to the spoken sentences) and a semantic component (which determines the semantic interpretation of a sentence). Chomsky then defines the deep structure as the structure which determines the semantic component, and the surface structure which determines the phonetic component. The surface structure of two sentences can be very similar and the deep structure very different. Consider the following:

I persuaded John to leave.

I expected John to leave.

In these cases, the surface structure (the way the sentences is presented, i.e. the order of the words, etc.) hides the underlying deep structure and therewith the structural differences of the sentences which are important to evaluate their meaning. This can easily be shown by transforming them into

I persuaded that John would leave.

I expected that John would leave.

Here, sentence (ii) is totally meaningful whereas (i) is not. The opposite is of course also possible: sentences similar in deep structure can differ greatly in surface structure, as can be seen in the transition from active to passive:

The doctor examined John.

John was examined by the doctor.

To account for the fact that all men can learn any human language as their first natural language, Chomsky claims that the deep structure is very similar in all languages. Moreover, it is also implied that humans cannot learn any language as a first language which does not obey the rules of a universal grammar.[28] The deep structure, so Chomsky, can not be represented in terms of associative networks.

Chomsky’s theory of grammar provides observable and unobservable grammatical features.[29] It is a matter of psychological experiments to verify if speakers’ judgments coincide with what is implied by theory. But nevertheless, Chomsky’s proposal of a universal grammar remains underdetermined. We cannot design any experiments proving the innate hypothesis. Also empiricists, so Harman, could explain the evidence in holding that inductive principles are biased in favor of transformational grammar.[30]

Harman objects that Chomsky’s theory, if taken literally, is implausible. He argues that the assumption that men already knew a language (a grammar) before they learned their first natural language ends in an infinite regress or a vicious circle.[31] This objection is certainly unjustified; Chomsky never claimed that the rules forming universal grammar determine a language. Instead, the child has to discover the grammar from the data that it is confronted with in experience. This happens in terms of hypothesis-making and verification.[32]

Chomsky characterized linguistic theory as a theory of competence (not performance) since it neglects the individual’s varying abilities and only focuses on the language faculties all men have in common. Harman accuses Chomsky to confuse performance and competence by confusing “knowing how” with “knowing that”. The speakers know how to use the rules of grammar, but they do not know what these rules are. Only the linguist can know them.[33] In particular, so Harman, Chomsky confuses knowing that certain sentences are grammatical unacceptable with knowing the rules that prohibit them.[34] He therefore concludes[35]: “Chomsky has not shown that the facts of transformational linguistics defeat an empiricist theory of learning.”[36]

The Status of Innate Principles

Principles as those on which a universal grammar might be based on are certainly not known by the average person (and even not by linguists since there is much controversy about the specific rules of a universal grammar). These principles are also not accessible by mere introspection. Nevertheless it is claimed that every person intuitively makes use of these principles. Likewise, many people have no explicit knowledge of the laws of logic, but still they are pretty good in applying these laws in everyday reasoning.

Chomsky admits that knowledge about grammar is often not explicitly known, just as Locke claimed for all truths in general (“for a man may live long, and at last in Ignorance of many truths.”)[37] Locke was simply wrong in stating that the Mind is capable in all it knows, since men often presuppose non-explicit knowledge in their actions. To put in a modern formula, knowing that p does not imply knowing that one knows p.

Chomsky heavily declines the view that language is just a habit or a “complex of dispositions to response”.[38] In what sense then are Chomsky’s innate principles dispositional? My proposal is that it is not the principles underlying universal grammar which are dispositional – they are in a non-propositional form actually there, represented in the mind -, but it is the beliefs unconsciously derived from these principles (which are beliefs about grammaticality) which are dispositional. In this view, only the principles of universal grammar are innate, not the beliefs.

What about the universality of such principles: is it sufficient to show that all human languages have a generative grammar in common? It was objected that the similarity of the grammar of human language could also be based on a common primitive language that all men had in common, ten thousands of years ago. This objection does not seem compelling to me and can certainly be refuted. Moreover, the fact that human languages differ so greatly in their appearance but not in their internal structure is great evidence for Chomsky’s hypothesis.

Locke rejected the view that innate principles are universal in another sense, namely that these are principles that everyone would consent to. First, he argues, that universality does not imply innateness, since there might be other reasons why everyone agrees upon a belief.[39] Second, there are simply no universal principles, since there are always people who deny even the most obvious principles (such as children and “ideots”).[40] Both arguments do not apply to Chomsky since the universality of universal grammar is not a universality of ideas or beliefs (and therefore not a definition of innateness) but a characteristics of all human languages. It is closely related with the faculty to learn a language at all, including the fact that almost every human is capable of learning any language as his first, since it is the universal structure of grammar that allows him this. Whether the beliefs about grammar generated by the underlying universal principles are universally consented to is a very different issue, and it is plain to see that this is not the case. Men even do not need to be aware by what their beliefs about grammar are produced, since people often are not aware of what causally produces their beliefs.[41] Because of this, also Locke’s confutation of the version of innateness as “that all Men know and assent to them, when they come to the use of Reason”[42] is not applicable to the innateness of grammatical rules. The use of reason might discover these principles, but often reason will not. In fact, even the linguists did not discover a distinct set of rules composing universal grammar (which is often held to be an empirical fact against universal grammar – Chomsky replied that these rules need not to be simple ones).

The principles underlying a universal grammar also do not have the status of the propositions in mathematics about which might be said, as Locke remarks[43], that the human mind has the disposition to discover them. The difference is that these linguistic rules correspond to a human faculty. Propositions in mathematics might be a priori, but they are certainly not innate (maybe with exception of certain logical truths). “Use of Reason” is simply not the issue in the innateness of grammatical rules (since we are concerned with the genetic issue). Similarly, Locke’s refutation of the version of innateness as “assented to as soon as proposed”[44] is mistaken.

The Computer Metaphor

Chomsky often talks about the human mind as an information-processing system, which receives sense data as input and beliefs as output. He then states that the output is richer of information than the input, and that therefore the human mind contributes something that is not contained in the input. This contribution is then due to innate principles. Stich objects that it is also the case in induction and abduction that the output is richer than the input.[45] This might be the case, but Chomsky gives good reasons why the contributions of the mind concerning language acquisition should be attributed an innate grammar. I do not think that Chomsky’s picture of the mind as (to play it up) a computer is just a metaphor, as was suggested by Adams. Instead he uses this picture to illustrate important mathematical features of his theory of syntax, giving profound descriptions of the mental processes that are at work in language acquisition. Chomsky’s picture, in my opinion, has to be taken very seriously.

Some Remarks about Concept Acquisition

The innatists only claimed that specific ideas were innate, such as the idea of God or the idea of truth.[46] Because of the interconnection of concepts with propositions, at least such ideas employed in innate principles have to be innate – how else could an innate principles otherwise be known? On the other hand, ideas cannot enter the mind by perception. In this sense, all ideas must be innate.[47]

Both notions of innate ideas are relevant in the nativist’s theory of concept acquisition. They attempt to give an account where our simple ideas come from (and in which sense they can be characterized as innate) and what mechanisms are involved when complex ideas are formed (i.e. when concepts are acquired). Adams claims that innate ideas then have to fulfill three conditions[48]: (i) men have the natural ability to form them, (ii) the exercise of this ability does not depend on any particular stimulation, and (iii) their formation does not depend on occurrences in our mind.

The controversy about concept acquisition between empiricists and nativists is even more unclear than it was already the case in the discussion about grammar. Both empiricists and nativists agree in that men have the natural ability to discriminate between simple ideas. They also both hold that children are born with the ability to combine simple ideas to complex ones[49] (as is e.g. also discussed in Locke’s Essay). Chomsky remarks that it is unclear whether the activity of the mind involved in concept acquisition is perception (i.e. selecting already distinct ideas on occasion of sense) or acquisition (i.e. making distinct what was confused before).[50] But, so Chomsky, also perceptual mechanisms must make use of an underlying system of generative rules.[51]

Fodor elaborated a theory of concept acquisition, proposing, in short, that concepts are triggered by experience and in this sense are innate (because not learned). In triggering, there is no rational relation involved between concepts and its causes. His proposals, however, seem to me rather ad hoc and not grounded on much evidence, so I will not discuss them here. Concept acquisition is an issue that still needs to be investigated by philosophers.

CONCLUSION

I have argued that Chomsky’s proposal of the innateness of grammatical rules, interpreted as faculties rather than dispositions, is a reasonable hypothesis which can withstand the Lockean objections. Chomsky’s view on language as a very specified faculty is also not vacuous since the language faculty is closely related to the intuitions about language, such as grammaticality, and is therefore indeed belief-generating, in opposite to other (e.g. non-mental) faculties.

On the other hand, there are still problems with Chomsky’s theory that are not addressed by Locke, but by contemporary philosophers:

· The particular internal factors and environmental factors involved in language acquisition are not determined yet, therefore we also do not know about how important each of them is.

· Chomsky claims that knowledge of language not a skill or habit. But there is some evidence suggesting that language use is an automatism (in fact, it is one of our best trained skills since we talk and listen a lot). On the other hand, holding beliefs is certainly not a skill. Therefore, some nontrivial problems about belief-production about language arise.

· Evidence indicates a more charitable view on human cognition that the innate hypothesis allows (as e.g. suggested by the human faculty to break codes, solve puzzles, etc.) So, if human mind performs great tasks in which innate principles are not involved, why should we go back to them concerning language acquisition? “That we learn what we learn is undoubtedly due in part to innate factors, but this does not go to show that what we learn is innate.”[52]

References:

Atherton, Margaret & Schwartz, Robert: “Linguistic Innateness and Its Evidence”

in: Stephen Stich (ed.), Innate Ideas, University of California Press, 1975

Adams, Robert M.: “Where Do Our Ideas Come From? – Descartes vs. Locke”

in: Stephen Stich (ed.), Innate Ideas, University of California Press, 1975

Chomsky, Noam: “Aspects of the Theory of Syntax”

The M.I.T. Press, 1965

Chomsky, Noam: “Recent Contributions to the Theory of Innate Ideas”

in: J.R. Searle (ed.), Philosophy of Language, Oxford University Press, 1971

Chomsky, Noam: “Cartesian Linguistics”

in: Stephen Stich (ed.), Innate Ideas, University of California Press, 1975

Chomsky, Noam: “Linguistics and Philosophy”

in: Stephen Stich (ed.), Innate Ideas, University of California Press, 1975

Cowie, Fiona: “What’s Within? Nativism Reconsidered”

Oxford University Press, 1999

Elman, Jeffrey L. (ed.): “Rethinking Innateness”

The M.I.T. Press, 1996

Goodman, Nelson: “The Epistemological Argument”

in: J.R. Searle (ed.), Philosophy of Language, Oxford University Press, 1971

Goldman, Alvin I.: “Innate Knowledge”

in: Stephen Stich (ed.), Innate Ideas, University of California Press, 1975

Harman, Gilbert H.: “Psychological Aspects of the Theory of Syntax”

in: Stephen Stich (ed.), Innate Ideas, University of California Press, 1975

Hart, W. D.: “Innate Ideas and A Priori Knowledge”

in: Stephen Stich (ed.), Innate Ideas, University of California Press, 1975

Katz, Jerrold J.: “Innate Ideas”

in: Stephen Stich (ed.), Innate Ideas, University of California Press, 1975

Locke, John: “An Essay Concerning Human Understanding”

P. H. Nidditch (ed.), Oxford University Press, 1975

Putnam, Hilary: “The ‘Innateness Hypothesis’ and Explanatory Models in Linguistics”

in: J.R. Searle (ed.), Philosophy of Language, Oxford University Press, 1971

Stich, Stephen: “Introduction: The Idea of Innateness”

in: Stephen Stich (ed.), Innate Ideas, University of California Press, 1975

Quine, W. V.: “Linguistics and Philosophy”

in: Stephen Stich (ed.), Innate Ideas, University of California Press, 1975

[1] e.g. the empiricist’s thesis that mathematical knowledge is grounded experience

[7] Stich points out that “innate” and “a priori” were often confused by rationalists, e.g. when Leibniz writes that “necessary truths are innate”.

[10] Cowie proposed that innate factors are represented by internal factors of the mind, whereas environmental factors are external. I do not think – as Cowie does – that the debate between nativists and empiricists is a debate between the relevance of internal and external factors and I find this identification mistaken, so I will put this issue aside. The factors producing innate principles may be both internal and external (genes are certainly external factors). And the statement that innate principles are mind-internal is obvious – what else could they be?

[13] Goodman, p. 143; I suppose that in this Dialogue the character Anticus propagates Goodman’s own views.

[14] Goodman is discussing a real puzzle: how can innate ideas be ideas without necessarily be used as ideas? We will see later in the chapter about concept acquisition how this puzzle might be resolved.

[20] This realism of the connectionist’ theory can be illustrated by a rough calculation: within the processing of a sensual stimulus approximately one hundred layers of neurons are involved. This can be shown in putting the reaction time of each neuron in relation to the overall processing time (which can be measured). On the other hand, estimates on how many nodes have to be involved in connectionist models yield a similar result.

[21] I have not discussed in what way an account for innate principles might differ from an account for innate ideas. Locke distinguishes between these issues in discussing them in distinct chapters (although he employs similar arguments to both). In fact both innate hypotheses, concerning principles and ideas, are closely related (holding a belief requires having concepts, and having concepts requires having beliefs about them). One of Locke’s arguments against innate principles therefore is that since there are no innate ideas, innate principles are impossible.

[24] I will interpret Chomsky according to what I wrote about innate principles as the principles underlying innate faculties. Therefore, I won’t characterize the beliefs produced by such principles itself as innate.

[28] But one can only learn such a language in terms of a puzzle, like a mathematical problem.

[32] Katz, p. 160; Atherton and Schwartz, p. 210, find the idea that the child makes decisions between hypotheses quite mystifying. It seems implausible that a child is able to make these decisions without reflecting on them. But if these decisions are automatisms, they do no really compare and choose among possible grammars. Chomsky writes about this issue that an explanatory linguistic theory also must entail a evaluation measure, leaving open the degree of consciousness accompanying such decision-making between concurring hypotheses.

[36] There is a (somewhat polemic) reply by Chomsky to Harman’s critique, but I fear that Chomsky

[40] Locke, E.I.ii.§4. Locke points out that even the logical principles “whatever is, is” and “it is impossible for the same thing to be and not to be” are not assented to by everyone, and if even these principles are not universal in this sense, then probably none.

[41] For example, participants were asked to compare the happiness of people’s faces on two photos, the same face which only differed in the (manipulated) size of the pupils. But nobody detected that it was the face with the larger pupil they attributed the greater happiness, but instead most participants stated that it was a (inexistent) bigger smile. So they were wrong about the cause of their (correct) judgment which was the happier one.

[46] For example, Descartes distinguishes innate ideas from adventitious and fictious ideas in the Meditations.

[47] Adams, p. 76; also compare Stich, p. 12, where innateness of ideas are characterized as an organisms property to be conditioned to stimuli from birth.