|« Language and Meaning: What is the Connection?||How Does Language Have Meaning? Part 3: Semantic Holism »|
How Does Language Have Meaning? Part 4: Language of Thought
How Does Language Have Meaning? Part 4: Language of Thought
This post continues the series on language and meaning. The first post asked the question of how language has meaning and how we can come to know what words and sentences mean. Since then, I have posted on a few different theories of language and meaning, most recently on semantic holism, which is the most popular theory within contemporary analytic philosophy. Today's post explains a novel approach to explaining how words, signs, phrases and expressions in language can come to have meaning – the idea that our thoughts are facilitated by an innate language and that the languages that we speak and write in are attempts to approximate our natural language of thought.
As was touched upon in the first post in this series, John Locke held that the meaning of words must be ideas. Each person has his or her own mind and the ideas in the mind must be unique to that person and they are also only accessible to that person. He argued that when a person speaks, that they first have an idea in mind and then try to find the appropriate words to express these ideas so as to allow others to understand these ideas. Likewise, when a person hears or reads words or sentences, they try to figure out what ideas the speaker had in mind when they said what they did. This idea-based semantics differs remarkably from all other semantic theories that were described in prior posts because none of these identifies the source of meaning with the ideas that someone has in mind when they communicate.
Jerry Fodor formulated a modern version of this, which he detailed in his books The Language of Thought (herein referred to as “LOT 1”) and its sequel LOT 2: The Language of Thought Revisited. Fodor argues that even beings that do not have language, such as animals, infant humans, and humans afflicted with aphasia, are nonetheless capable of thinking. From this Fodor concludes that thinking cannot be completely dependent on any inter-human communicative language (LOT 1, p. 57).
Fodor's theory is based in part on the computational theory of the mind and the related representational theory of the mind. Fodor analyzes how thinking must work and points out that the brain is functionally representational and computational and computation necessarily involves the manipulation of symbols. From this, he concludes that at least some mental processes are computations on mental representations (LOT 2, p. 20). Fodor argues that the symbols that are being manipulated inside the brain and which represent ideas that correspond to things in the world constitute an inner language. According to Fodor, this inner language cannot be the same as any so-called natural language such as English or Chinese because it is possible, to a large extent, to translate a text from one inter-human communicative language to another while preserving the ideas that the text is about.
For example, if a native English speaker were to read and understand something written in English and then a native Chinese speaker were to read and understand the same passage after it is translated as accurately as possible into his own language, then the result of this will probably be that the two will have similar thoughts, and these thoughts transcend any inter-human communicative language. So it is not that an English speaker thinks in English and a Chinese speaker thinks in Chinese, but that there is a certain language that exists inside each person's brain that facilitates pure thoughts.
Fodor argues that this language of thought, which can be called mentalese, exists in one form or another within the mind of any being that is capable of thinking. Mentalese is thought to be more natural than any so-called natural language because it is supposed to be innate to the brain, whereas any language used for inter-human communication was created by humans. Mentalese is thought to have an indirect relationship to inter-human communicative languages in that a person thinks in mentalese but figures out the best words to use for inter-human communication and then speaks or writes these words.
According to Fodor, there are features of mentalese that are similar to inter-human communicative languages. For one, he argues that thoughts are fundamentally comprised of certain atomic symbols that represent certain basic concepts and that more complex concepts are formulated through some composition of these basic concepts. He also argues that smaller units, mostly words, lexemes, morphemes, are the basic units of semantics in thought and that sentences are constructions of this. One important point of argument that he has advanced is that these atomic symbols in the mind are combined using syntactic rules that are likely analogous, at least to a certain extent, to the syntactic rules that are features of inter-human communicative languages. As Fodor explains:
One of the respects in which Mentalese is supposed to be language-like is that its formulas exhibit constituent structure. To a first approximation, the constituents of a discursive representation are its semantically interpretable parts, and it is characteristic of discursive representations that not all of their parts need be semantically interpretable. Consider a complex expression like sentence (I) below. It's constituents consist of: the sentence itself, together with the lexical items 'John', 'loves', and 'Mary' and the two phrases 'John [noun phrase]' and '(loves Mary [verb phrase])'. By contrast, it has among its 'parts' non-constituents like (e.g.) 'John loves' and (assuming that 'parts' can be discontinuous) 'John … Mary'. So, then, every constituent of a discursive representation is one of its parts, but not vice versa.
(I) John loves Mary.
I assume that all this holds for mental representations too. As usual, the arguments in the case of Mentalese run parallel to the arguments in the case of English. (LOT 2, p. 59)
Among those who favor this theory, there is a lot of disagreement regarding which concepts are fundamental. Fodor has argued that nearly all words in one's language that are not compound words are innate atomic concepts in the brain, including words like “tree”, “chair” “carburetor”, “horse” and the like (LOT 2, p. 129). He believes that even the word “kill” has an atomic equivalent in the mind, rather than this concept being a mental construction of the simpler concepts “cause” and “die”. On this note, Fodor's opinion is quite eccentric. Stephen Pinker has suggested a more moderate version of mentalese in which complex concepts like a carburetor or an umbrella are constructed from simpler concepts within the mind and where the only atomic concepts are those that couldn't conceivably be understood as a construction of any simpler concepts (The Stuff of Thought, p. 97).
A counterargument to the idea of a private language comes from Ludwig Wittgenstein, wherein he seeks to show that this sort of account is incoherent. Wittgenstein defines a private language as one that is understandable to only one person and cannot conceivably be understood by anyone else. It would seem that something like mentalese, as Fodor and Pinker have described it, would qualify as such because it is impossible for someone to experience someone else's thoughts, and thus it follows that it must be inconceivable that these thoughts could be understandable to anyone but the one who thinks them.
Wittgenstein addresses whether there could be any internal experience that the word “pain” refers to:
If I say of myself that it is only from my own case that I know what the word “pain” means––must I not say the same of other people too? And how can I generalize the one case so irresponsibly?
Now someone tells me that he knows what pain is only from his own case!–––Suppose everyone had a box with something in it: we call it a “beetle”. No one can look into anyone else's box, and everyone says he knows what a beetle is only by looking at his beetle.––Here it would be quite possible for everyone to have something different in his box. One might even imagine, such a thing constantly changing.––But suppose the word “beetle” had a use in these people's language?––If so it would not be used as the name of a thing. The thing in the box has no place in the language-game at all; not even as a something; for the box might not even be empty.––No, one can 'divide through' by the thing in the box; it cancels out, whatever it is. (Philosophical Investigations, §293)
In this, Wittgenstein is arguing that any private concept, be it the first person experience of pain or the atoms of meaning from which the complex concept “carburetor” is constructed in the mind, if they exist are like objects in boxes that we all have but cannot speak about because they are fundamentally private to us and therefore they cannot conceivably be made public. Just as when everyone speaks of this “beetle” in their box, it might be different for each person and it might be changing constantly, so is the situation of trying to talk about the contents of one's own first person experience. Each person has their own experience and it is not possible to have someone else's experience. Wittgenstein is arguing that while there is a word “pain” and we can talk as if we have a shared understanding of what this is, that it is not possible for us to have this shared understanding.
For Wittgenstein, thought is inevitably tied to language, which is inherently social. Therefore, there is no 'inner' space in which thoughts can occur. Wittgenstein believed that one's public language is their language of thought. So an English speaker thinks in English and a Chinese speaker thinks in Chinese. This conception of how a person thinks leads to the strong version of the Sapir-Whorf hypothesis, in which one's language determines their possible thoughts and puts certain constraints on their cognitive capabilities.
Who is correct here, Fodor or Wittgenstein? What theory of language and meaning makes the most sense? Is it mentalese, as explained here, or perhaps is it semantic holism or logical atomism or some other theory? In the next few posts, I will provide what I believe are the best answers to these questions and my reasons for believing as I do.