Science Without Language
Clearly, there could not be poems or novels or essays without language: these things consist of words arranged into sentences. Equally clearly, there could be athletic activities without language: football, high jumping, sprinting, badminton. People may talk as they engage in these activities, but they don’t consist in talking. Could there be art (painting, music) without language, or economic activity, or politics, or carpentry? Some may say no, but that is certainly not obvious (I don’t think these activities are necessarily language-dependent). Could there be science without language? That question is much murkier; it is like the question of whether there can be mathematics without language, or logic, or geography, or history. If we ask whether there can be science without thought (knowledge, belief), the answer is quickly returned: no, because science isscientific thought. Science can’t exist without minds for it to exist in, and propositional attitudes are its necessary vehicle. There is no such thing as scientific knowledge if there is no scientific belief. The sciences, as we have them, are bodies of belief or knowledge, so there can’t be the former without the latter.[1] In the same way, there can’t be science (as we have it) without mathematics or logic or observation or inferential reasoning. Viewed as a human cognitive structure, science consists of all these things, but centrally of thought: there is no such thing as thoughtless science. Then the question is whether scientific thought requires scientific language. Can there be science without symbols? Someone might say that there cannot because allthought requires language. Such a person might mean an internal language or an external language—a language of thought or a language of speech. The former view would imply the scientific necessity of language trivially; the latter view would imply it by making inner thought depend upon outer spoken language. I am concerned with the second question: so, we are asking whether science depends on speech. I don’t think all thought depends on speech, for reasons I won’t go into here; I am interested in the question of whether scientific thought in particular depends on speech. Is there anything specific to science that makes it essentially linguistic? Is it like poetry or is it like painting? Is the presence of spoken language necessary or contingent to the existence of scientific thought?
It might be replied, evidently plausibly, that science, as we have it, does essentially involve language because of the existence of scientific communication: conferences, journals, conversations, books, letters, peer review. But that answer is superficial: why shouldn’t there be a Robinson Crusoe figure keenly interested in science and yet cut off from all scientific communication with other scientists? He has scientific thoughts (observational and theoretical) but he never talks about them with anyone. That seems perfectly possible; and isn’t the ordinary scientist in essentially this position when alone in his lab or study? Talking about science is subsequent to thinking about it not a pre-condition of it. Any serious connection between science and language must cut deeper than this. It can’t be just that speech is the externalization of scientific thoughts, if the connection is to be of any significance. Art requires artistic materials in order that anything be made (e.g., pigments and sounds); is there anything about science that makes it require linguistic materials? Does it require, say, the existence of grammar in order to count as science? Granted, not all thought presupposes grammar (say, animal thought), but does scientific thought need the resources provided by grammar? The question is suggestive and appropriate, but the answer to it appears plainly in the negative. There could be science without syntax. What if we had evolved to the present time with the general intelligence we now possess but without ever acquiring spoken language? Language came along recently in human evolution, but our big brains were there all along; so, in principle, it looks possible for us to have developed scientific thought but not spoken about it (solitary Newtons and Darwins, say). And yet it seems funny to say that our present scientific knowledge owes nothing to language, though it may be difficult to identify what it is exactly. Our intuitions are pulling us in two directions: on the one hand, thought as such does not entail spoken language to express it; on the other hand, our scientific world-view seems steeped in language, and inconceivable without it.
The science of linguistics needs language, obviously, because it is about language. But most science is not about language but about the extralinguistic world. We can truthfully say that science requires more than just scientific beliefs: it requires observation, memory, and theory construction—but where does speech come in? Compare history and geography: they too require observation and memory, possibly also theory construction, but do they also require language? Well, look at a typical history or geography book—what do you see? You see sentences, dates, maps, but also names—names of people, names of places, names of movements. If you were to delete these names, you would be left with very little. The knowledge you acquire from these books is typically name-involving (Paris is the capital of France, say). True, you could have some historical or geographical knowledge without the introduction and use of names, but those subjects would be crippled without the apparatus of naming. Such knowledge is name-centric. Thus, if we ask whether history or geography requires language, the answer is yes—part of language, at least. You couldn’t be a languageless being and have our geographical knowledge, because that requires mastery of the practice of naming. Geography would never have got off the ground without naming as a pre-existing psycholinguistic achievement. It could exist in embryonic form but not in its current splendor. Huge amounts of history and geography are about names, in the sense that you have to be aware of what is involved in something being called by a certain name: these subjects are implicitly metalinguistic. One knows, for example, that the city called “Paris” is in the country called “France”. So, they resemble linguistics. Not every concept is like this: some concepts, and the thoughts they feature in, are not tacitly metalinguistic—color concepts, shape concepts, moral concepts, etc. But concepts like Paris and France are, so knowledge involving them is language-dependent. To be Paris is to be called “Paris” (but to be red or square is not to be called “red” or “square”).
The extension to science is obvious. Enormous tracts of scientific discourse consist of name-like expressions–labels, tags, designators, cognomens–and hence introduce a metalinguistic element. You can’t grasp the propositions expressed without understanding the practice of naming. You have to be name competent, and hence a speaker. Without the use of names science would be crippled: just consider zoology and astronomy, to name but two sciences. The ability to name things, often using what are called “technical terms”, is critical to advanced science; that is why neologism is so common in science. The roots of naming no doubt trace back to the vernacular, but this resource is massively exploited in the sciences; and it is very useful there, because we often don’t know the nature of the things we wish to refer to—we need a nondescriptive label. If we could replace all names in scientific discourse with general descriptions, we could in principle dispense with language as an aid to scientific thought; but in practice that is impossible, so we are stuck with them. The result is that science cannot do without language, at least for limited beings such as ourselves. It can exist at a primitive level without language, but once we start to insert names into our scientific statements, we are introducing language into scientific thought. The answer to our question then is that much of science is language-dependent, though not all. Science as we have it requires language mastery for its possession.
This changes our picture of scientific theories. Empiricism pictures science as a congeries of experiences not essentially bound up with language—observations (not observation statements). That is the cognitive kernel of our scientific knowledge. But, according to the position here advanced, it is also a linguistic construction—a congeries of names (name-like expressions). Nor are theories “sets of propositions” that may be grasped by the nonlinguistic, but assemblages of words accessible only to the linguistically initiated. This is “nominalism” not empiricism (or even “propositionalism”).[2] Our scientific knowledge is constrained by our cognitive capacities (trivially), but these involve the apparatus of naming with its distinctive features. Scientists thus need to be speakers in order to do science in any meaningful way. Thinking isn’t enough (though necessary). This is why you have to learn the nomenclature of a science in order to do that science. It is the basis of categories and classification. A monograph entitled “Naming and Science” would be well named. Fortunately, we are prodigious name-users in ordinary life, so the cognitive demands of science are not too daunting. In all probability, we evolved this capacity as a tool of social interaction—we needed names for other people. So, the basis of scientific language is social psycholinguistics (in part). No human science without human society. We called each other by names and we then extended that practice to the rest of the universe (notice how personalized astronomical names are). Our intuition was therefore correct when it prompted us to declare science (partly) parasitic on language—without endorsing the strong and implausible claim that thought is always dependent on language. An adequate theory of scientific knowledge must include an account of scientific naming (a neglected field).[3]
[1] It will be noticed that I bypass completely such views as that scientific knowledge is a “language game” or a series of “texts”. That is an impoverished and misleading picture of what science is. In the first instance, scientific knowledge is a type of psychological state (competence, cognitive structure). So, there is no quick route from science to language.
[2] I don’t mean “nominalism” in its usual acceptation; my neologism is intended to capture the idea that scientific knowledge partly concerns what things are called, i.e., is metalinguistic.
[3] I often think that philosophy of science covers a too narrow range of topics. It has been overly influenced by logical positivism. This paper attempts to introduce a new topic into the subject.
