By an idea from Wilhelm Humboldt in the early 1800s, languages vary in how they ‘Glue’ words together. English does this only in a quite marginal way, varying from one variety to another, as in English English wouldn’t from would not. Going beyond Humboldt, this glueing allows one glued structure to be glued to another, with languages varying in terms of what can be glued to what, how strong the glue is, and how far the elements are modified in the process. All of this obviously falls within the space of what the child has to learn.

Putting things together

In a novel way, I hypothesise that ‘glueing’ applies to the sound system or ‘phonology’, as well as to the ‘syntax’ – assembling words and parts of words together into structures with well-defined meanings by what is often known as the ‘grammar’. This application to phonology must logically have preceded its application to syntax. There had to be speech sounds or phonemes and words, or there would have been nothing to glue.

From a given evolutionary point speakers started to Glue available acoustic elements together, clicks, grunts, sighs, moans, hugely increasing the vocabulary or ‘lexicon’, by the mere fact that the elements were cognitively organised. New ideas with no obvious ostensive expression, could be expressed by arbitrary and abstract combinations of elements, glued together in a sequence of steps of gradually increasing complexity.

By the hypothesis here, Glue applies to all structures, including phonemes, syllables, and words, even though some instances of the sequencing are hard to perceive.

Gluing for glory

By an evolutionary process, by glueing glued structures to one another, this cognitive architecture made it possible to Glue, not just words, but simple sentences:

• Bird die

• Dad kill bird

• Dad kill bird in wood

• Think Dad kill bird in wood

The same elements could then be decomposed, giving speech sounds or phonemes which could be glued to give syllables and make did, dead, dud, dod, dood, all possible words, allowing the lexicon to expand exponentially.

By my hypothesis here, all languages use Glue in their phonologies, to form the phonemes from more basic elements or ‘features‘ and going beyond this to assemble the phonemes into words.

The architecture could be reapplied to Glue syllables to form succesively longer words, increasing the possible lexicon by another exponent for every act of Glue.

There is no reason for assuming that this evolved other than over a historical period, a number of generations.

On this understanding, it is only by Glue that it is possible to state facts which are either true or untrue.

The most basic elements

In the study of speech and language there an obvious tension between the externality of what is heard as a complexly modulated airstream and what is formally organised as a series of nested structures. On one approach, the two are quite separate. On another they are part of one system. Noam Chomsky and his 1968 co-author, Morris Halle, have gone decisively from the former to the latter. I propose that the evidence of children’s speech and language and the conceptual necessities of evolution decisively support the former.

Not the last step

By this hypothesis, Glue can’t have been the final step in human-specific linguistic evolution. It allows only a small fraction of modern language. It does not allow elements within the sentence to be linked together other than by its own simple formalism. There must, I believe, at least one later step. By one later step, one or more of the glued elements could be reworked. I propose that the key driver of this subsequent evolution was to do something, in a way otherwise impossible.