28 Apr 2012

On the computational nature of syntax

I found an amazing article called On the nature of syntax (2008) by Alona Soschen who, in a nutshell, uses language as a means to examine possible underlying features common to other adaptive systems. Strangely enough, this intrigues me as a programmer too. To quote the abstract:
"There is a tendency in science to proceed from descriptive methods towards an adequate explanatory theory and then move beyond its conclusions. Our purpose is to discover the concepts of computational efficiency in natural language that exclude redundancy, and to investigate how these relate to more general principles. By developing the idea that linguistic structures possess the features of other biological systems this article focuses on the third factor that enters into the growth of language in the individual. It is suggested that the core principles of grammar can be observed in nature itself."
While this is a powerful subject in itself, there are also some interesting facts mentioned within about extreme language structures. It's stated that nouns technically rank higher than verbs universally speaking and this helps explain why the Australian language of Jingulu quite astonishingly has only three true verbs in its vocabulary: "do", "go" and "come". More extreme yet, the Nigerian language Igbo (aka Ibo) has, in place of verbs proper, inherent complement verbs which are made up of -gbá plus a noun (eg. –gbá egwú "to dance", literally "do dance"; –gbá igwè "ride a bicycle", literally "do bicycle"). This confirms my prior impression through my experience with computer programming that verbs are equivalent to "computer functions" that operate on input data (ie. nouns). Nouns then indeed are most primal since one must have data first before any function can operate on it. Nonetheless verbs too are a close second in importance since not much could be expressed without them and likewise not much could be programmed without functions. The interrelationships between language, logic and the qualities that create an adaptive system keeps me busy for hours.

14 Apr 2012

The first person pronoun in Afro-Asiatic languages

My mind lately has been seduced by some non-linguistic, programming-related material I've been researching busily on the side. However I'm ready to hop back into some lingering issues in my comment box where the latest discussion had ended off with the dilemma of reconstructing the first person pronoun in Proto-Berber. Let's zone in first on this Proto-Berber word for "I" before extending the topic to the rest of the Afro-Asiatic family.

Among available Proto-Berber reconstructions, my personal notes so far record Kossmann's *nǎḱḱ and Dolgopolsky's *ənakkʷ. However to compensate for Zenaga's cognate niˀkan with an out-of-the-blue glottal stop, and in keeping with known sound correspondences, alternatives like *nəʔḱḱ and *năɣḱ are suggested by my commenters. Afterall, if both Proto-Berber and produce a Zenaga glottal stop, these are reasonable ideas on the surface. However I've expressed my dismay about these word-finals because I don't find them to be terribly "pronounceable". This is my plebeian shorthand for saying that the forms in question border on the phonemically exotic at the expense of phonotactic rules.

I figure that if we're going to brainstorm it would be best to keep one's fancies to some clear-minded structure that includes a vision of not only the sound inventory of the protolanguage but rules governing its syllable structure. In both alternatives, we have a rather heavy word-final dump loaded with a bunch of articulatory features that, according to general rule, should be the least able to maintain complex contrasts in world languages because of the issue of saliency in an expiratory position. This to me just raises more questions than answers. In a form like *nəʔḱḱ, we would have an inaudible glottal stop nestled within the shadows of a palatalized geminate. To reconstruct word-final *-ˀḱḱ is to imply that there are simpler codae that contrast with it such as *-ḱḱ or ˀḱ but I suspect that such exemplary forms will be hard to come by leading to a reasonable suspicion against the validity of such a sequence. Likewise, if we go with *năɣḱ, we're left with similar questions. Where then are the examples of its implicit voiceless counterpart *-xḱ? How is this alleged word-final palatalization to be articulated? Would this palatalized velar plosive be pronounced with pre-palatalization or would it be accompanied by a non-phonemic release by way of palatalized aspiration or a subtle vowel? All of these issues must be addressed and it's not sufficient to me to only pay heed to phonemics. What is unavoidably enmeshed with the riddle of its exact form is simultaneously the manner in which these phonemes may be assembled into valid Proto-Berber words. We must pay attention to both issues at once in order to provide more realistic alternatives. These questions are just as much questions that *I* must address, and do indeed feel obligated to address, when seeking a better answer.

After much deliberation I realized one piquant possibility that agrees with my determined obeisance to Occam's Razor while (hopefully) being more congruent to extent facts. Abandoning my previous skepticism for a moment, let's simply accept for the sake of argument that some palatalizing element is original to Proto-Berber because of pesky details like Shenwa's cognate nəč. Why then not opt for a more streamlined form like *nəky? In this way, the glide can credibly serve as both a source for palatalization and gemination at once. It satisfies my phonotactic constraints which restrain me from indulging in overloaded coda of more than two consonants and, if this may be pursued to its ultimate conclusion, it might reduce Berber's commonly reconstructed phonemic inventory by eliminating palatalization as a phonemic feature altogether. Note too that as I write this, I recognize that I must investigate how Zenaga's -ˀk- in its cited pronoun niˀkan contrasts with geminated kk because if it doesn't and if it's merely a matter of orthographic style, then there should be no issue left regarding the possibility of such a glide creating later geminates in light of the fact that in Three irregular Berber verbs: 'eat', 'drink', 'be cooked, ripen' (2008) Maarten Kossmann posits quite similar developments for Proto-Berber, or at least for "Pre-Proto-Berber".

Yet this still leaves me with a mysterious remainder to solve: How does this Berber pronoun relate to the rest of Afro-Asiatic (AA) and its forms? I have doubts towards the notion that the final palatalizing element, whatever its nature, is original to Afro-Asiatic. While we have Classical Hebrew אָנֹכִי (ˀanoki) alongside Akkadian anāku, the former is likely contaminated with the 1ps possessive suffix. This implies the latter to be more conservative, thus Semitic *ˀanāku with final *-u. In Egyptian too, evidence mounts against a final *-i or *-y when we observe that there is no palatalization evident in the final velar of ỉnk (> Sahidic Coptic anok) in contrast to the 2ps feminine dependent pronoun with palatalization: ṯm (*cim) < AA *kim. This leads me to wonder if Berber, like Classical Hebrew, has innovated by contamination with the same common 1ps possessive element *ya found throughout AA.

6 Apr 2012

Emergence of the Rank-5 society

The evolution of cognition by William Benzon and David Hays is an endlessly fascinating read. I get the same sort of inspired buzz as when watching the Matrix and probably for the same reasons.

Their basic proposal is that human societies can be classified according to different ranks representing different modes of thought as we edge towards more complex societies. As an overview, they explain that Rank 1 is associated with the invention of language, Rank 2 with the invention of writing, Rank 3 with the invention of calculation, and Rank 4 with the invention of computation. Each stage of that procession, they explain, is dominated by a certain way of seeing the world that adds something new and valuable to our collective understanding in the previous stages. It has as much to say in sociology as it does in the science of computation.

I take away a lot of new ideas and questions in this piece. One curious absence in the entire article is a direct mention of a Rank-5 society. What would that entail? What would its hallmark invention be? I come to the conclusion that it's a society that through the medium of machine language has delegated the process of algorithm creation to digital agents, through the process of universal induction and by a mechanism of conscious adaptive system design.

In such an age, I gather that beyond our need to "control" systems, as now, the new way of seeing the world will recognize that a means to balance is paramount in all lasting systems. The notion of "control" thus will evolve to a point where we accept a hands-off approach by creating a good system to begin with that suits our needs, a system that no longer requires our direct involvement because its embedded balance keeps it dependable. Self-managing systems will become the norm, breeding a whole new way of seeing the world and our place in it. The beginning of this age then will be announced by the emergence of AI.