Nativism & Empiricism

Here's a quick introduction to an ongoing controversy about the operational definition of language. We'll get down to it by asking two simple questions: 1) Where is language located? 2) Where does language come from?

Biology, genetics, psychology, neurology, artificial intelligence, anthropology and so many other fields have contributed to this debate. Great minds have even hung their hat on one particular answer to those two questions. In the abstract, they have offered two main answers.

The first is linguistic nativism, which holds that the basics of language and grammar are innate. This is the "nature" or "internal" perspective. If these things are innate, they are built into every human being at birth. It follows that, for nativists, the origin of language is rooted in the history and development of the biological organism which contains that language instinct. They point to apparent linguistic universals shared by all humans as evidence. For them, language is in the genes.

The second perspective is linguistic empiricism. Empiricists think that language is entirely learned. This is the "nurture" or "external" perspective. If language and grammar are learned, they are features of the organism's environment. Empiricists point out that we lack evidence for language-specific adaptations in human biology and that all known features of language match features of other cultural developments. For them, language is a cultural artifact.

Other options abound. For instance, some cognitive psychologists think that the brain has essential equipment for making sense of grammar (against the strong empiricist assertion), even if this equipment contains no linguistic information like grammatical categories or frameworks for pronunciation (against the strong nativist assertion). But we can then ask a subquestion: is this equipment specific to language processing or more general in its abilities? This question reopens something like the nativist/empiricist split at the cognitive level.

These perspectives influence the way proponents collect and evaluate linguistic data. Linguistic data are sounds and text gathered in specific languages in order to study those languages.

Nativists are likely to think that, since people are born with a universal grammar, we all have an intuitive grammatical sense. Speakers can judge a grammatical sentence from an ungrammatical one. For the typical nativist, you check if a sentence is grammatical by asking a native speaker. Against this, empiricists charge that actual use (not intuition) determines which sentences occur and which do not.

For the average empiricist, you check your sentence against a corpus containing many examples of the language you're asking about. The more frequent its appearance, the more likely the sentence.

The perspectives I just outlined don't offer equal footing to their adherents, not even given the limited evidence currently available. "Strong" nativism has seen a sharp decline among non-linguists and a slow but steady retreat within linguistic circles. But that doesn't mean everyone's championing the tabula rasa, either.

The topics in my linguistics lessons for language learners start from less controversial claims. First, the human body makes noises through the air and marks on the page that other humans recognize as language. Second, a variety of languages have been used in a variety of ways in a variety of societies throughout human history. We observe difference (change) over time and space. Third, concrete observations of people using language ground abstractions made based on that data, and not the other way around.

  1. Humans use their bodies to speak, listen, write and read. (Definition.)
  2. What is spoken, listened to, written and read varies over time and space. (Comparison & examples.)
  3. We explain language based on what we observe of speech, listening, writing and reading. (Observing & theorizing.)

Crucially, these claims are readily available to the average educated person to check, while the nativist/empiricist divide would ask lay readers to engage in speculation without strong support, and so the null hypothesis remains in effect however enticing the claims. (Although note that the null hypothesis "language is not an internal biological feature of humans" is tantamount to the weak empiricist claim.)

The "nativ(e)" of nativlang.com and Native Grammar is a nod to the concept of learning languages well, not the nativist/empiricist debate. The lessons on this site go on to consider the basics of how language works, straddling all three of the above claims by including support for and implications of each claim.

I hope you've gained some understanding of this meta issue. It will allow you to appreciate the tension between variations and universals in discussions about human language.