Pre-Human Language

(This page is a work in progress.  To the best of my knowledge, the ideas I present here are uniquely my own.  This is based on an idea I first put in writing on 11/24/2011 when I was learning about noun incorporation in Takelma.  I last updated this page on 9/3/2013.)

Human language production and understanding is complex, and involves the coordination of several parts of the brain.  I find it difficult to believe that this faculty evolved all at once, and much more likely that it evolved in stages, gradually co-opting parts of the brain that had originally had other functions, making the final product appear somewhat cobbled together.  In this page, I'll make a proposal about how I think pre-human language may have evolved.

My proposal aims to use an evolutionary narrative to explain two things about human language.  First, I want to propose a gradual process by which the language centers of the brain could have come online one by one, rather than evolving to a state of completion all at once.  Second, I want to suggest a reason for the different but overlapping capabilities of analytic and synthetic processes across human languages.

For the purpose of this argument, language is the use of sounds as abstract symbols to communicate information.  Pre-humans no doubt vocalized in many important ways, but at some point discrete abstract sounds came to be symbolic of discrete ideas, and that is when language began.  The degree to which language could convey high-value information was the degree to which it contributed to the survival of the group, from which point forward language and evolution became intertwined.

Phase I: Monomorphemic Language

I propose that language processing began with the Middle Temporal Gyrus, where lexical semantic processing is carried out now, and consisted of single utterances with culturally-assigned meaning, but no internal morphological structure.  In this phase there would have been no real distinction between morphemes and words, and there was no significant working memory to build phrases.

The earliest speech acts may have served to draw the attention of the group to something in the immediate environment, thereby passing on valuable information.  Calls like this could draw attention to dangers (lion!) as well as resources (berries!). The referent of an attention call is an object, making the attention call like a noun. Attention calls are useful as long as the group understands the importance of the object referred to.

However, imperatives (run! gather!) are even more useful to organize the behavior of the group in the face of new and more complex situations.  Imperatives are like verbs, so the differentiation between attention calls and imperatives may have been the first differentiation between parts of speech, and ancestral to the noun/verb distinction.

Intonation is a strategy that can be applied to both nouns and verbs to create different types of speech acts without calling on morphology or syntax.  I suggest that speech acts like questions (lion? run?) could have appeared in this phase of development as well.

Phase II: Paleosynthetic Language

I propose that the next major phase of language development involved the Superior Temporal Gyrus, and with this came the introduction of the predecessor of syntax, which I'll call paleosynthesis. Paleosythesis allowed speakers to relate monomorphemes together in a larger structure, like a phrase, but using a strictly limited amount of working memory, and with no ability to nest phrases within each other.

In this phase, nouns and verbs became more clearly distinct categories, and they could be joined together in rigid paleosynthetic phrases.  Words and morphemes would have still been essentially the same thing, with no internal morphological structure, but the paleosynthetic phrase acted as a sentence.

I hold that the strategies developed in this phase are ancestral to the synthetic strategies used in modern human languages.  For this reason, I think it is likely that markers of person, tense, modality, aspect, number, and gender also evolved at this point.

At the end of this phase of development, speakers would have been able to formulate fairly complex ideas, like "The fish are many.  We will go catch them."  But long narratives would have to be broken down into many short declarations to be processed and understood in series.

Phase III: Human Language

I propose that human language arose when the Inferior Frontal Gyrus was co-opted for syntactic processing.  Paleosynthetic phrases could now be organized into larger, more flexible structures, and with greater working memory, recursive structures could be built.

The paleosynthetic phrases of Phase II became the words of Phase III, and paleosynthetic grammar developed into morphology, but retained its rigid structure and lack of recursion.

The new syntactic grammar is completely capable of expressing everything that the old paleosynthetic grammar could express, and more.  In modern language, however, both strategies coexist, and often overlap.

For example, the following verbal complex in Takelma could reflect the paleosynthetic strategy held over from Phase II.  In Takelma, a word like this would have been used as one component of a larger syntactic structure, but at the same time it contains within itself everything a sentence needs, so it could also stand on its own.

han-waya-swilswàlhi
He tore him open with a knife.

This complex contains a verb (swilswàl), implied subject and object (he, him) and an instrument (waya, "knife").  But this is not a verb phrase in the generative grammar sense, and waya is not a noun phrase.  You may not add adjectives or modifying clauses to waya while it is embedded here--to do that, you must move it out of the verbal complex and replace it with a coreferent pronoun.

Under this explanation, inflected words in modern languages are like frozen islands of ancient linguistic strategies, often duplicating the work done by the more fluid, modern strategies that were built around them.

As an initial stab at this question, I am probably wrong on a number of points, but I think my general argument holds water:  Language must have evolved by stages, and modern language probably retains evidence of the course of that evolution.  I think the distinction between synthetic and analytic strategies is one likely artifact of the course that linguistic evolution took.