5.1 Introduction

It is widely accepted in the scientific community that Language is produced by an organism, which via its motor cortex, sets in motion vocal and hand-movements. Recent FMR studies show motor cortex activity for verbal motor tasks (Brown & Hagoort 2003:220). However, little is known about the specific biochemical mechanisms the brain activates or inhibits when using language. One of the most important discoveries of the XX century has been to realise that only approximately 30,000 genes can generate 20 billion neurons in a human brain and produce remarkably varied behaviour (Marcus 2004). Despite this neuronal explosion, it is revealed from imaging studies that only a small fraction of these neurons are activated for language production or interpretation and that regardless of the apparent complexity that any language may show, a human brain can effortlessly master it with little exposure to the environment. This fact may perhaps indicate that few genes and little environmental exposure may be necessary and sufficient to produce much variation in behaviour. What seems universally accepted by followers of neutral monism, is that knowing how gene regulation affects the physiology and biochemistry of brain cells, will eventually elucidate how linguistic effects emerge from these anatomical and biochemical interactions (Edelman & Tononi 2001, Marcus 2004).

In view of this few-genes-not-much-exposure fact, I will take the simplest possible hypothesis to approach variation: an organism might use the same operative mechanisms, at different levels of organization producing variation, unless it is proved that it does not. This from-general-to-particular step by step approach is compatible with the hypothesis that two allegedly different species or organisms do not start from scratch in their developmental differences. It is also coherent with the fact that the same operative mechanism may create big different effects depending onto what, where and when it operates.

Following on this simplified approach, it is my intention here to bring back into the research programme a working hypothesis promoted originally by N.Chomsky in the 60s (Chomsky 1965, Chomsky & Halle 1968). His initial hypothesis was that the phonological level of organization: [p+a], here segment sequencing, and the syntactic level: [John + walks], here word sequencing, could be operating in similar ways. Under this simplification hypothesis, the "mapping" from an underlying level to a surface level (utterance) was carried out by varied types of linguistic operations, called then transformations. This unification working hypothesis was abandoned later on, concentrating instead on the mechanisms needed to switch word order within a hierarchical syntax (movement transformation), and its mapping into logical form. Phonology, like meaning, were considered secondary independent systems. More recently, Chomsky has reduced the Faculty of Language to an even more isolated computational system which generates discrete infinity and maps its outputs onto two other systems by interfaces: "a computational system (narrow syntax) that generates internal representations and maps them into the sensory-motor by the phonological system, and into the conceptual-intentional interface by the semantic system" (Hauser,Chomsky and Fitch 2002:1571).

(1)	Broad Faculty of Language (LFB)

Some specific data taken from Spanish and other languages will be discussed. The analysis suggested here could be interpreted as not supporting the claim that there is a core FLN (narrow syntax) which maps onto Phonology. The data discussed here may support instead the claim that Phonology is not peripheral as opposed to the core nature of narrow Syntax, and that therefore this dichotomy seems to be unjustified. The evidence comes from the finding that segmental sequencing (Phonology) seems to share with word sequencing (Syntax) common integrative effects discussed in detail below. Moreover there is evidence that one form of integration, here called stereo integration, is common not only to phonology and syntax (Garcia-Bellido 2000, 2003) but also to the visual perceptive system(Zigmond et al. 1999:ch.13, Kay 1991). In (2) below there is a visual representation which can be integrated as containing one bottle and two cups, one at each side. The phonological representation in (2), is integrated as containing a Consonant and two Vowels, one at each side. The syntactic representation in (2) is integrated as containing a Verb and two Nouns, one at each side. This symmetric representation will be referred here as Stereo. Three elements may be integrated in a spatial grouping, as shown in the visual representation of (2), or in a spatial (written) and temporal (spoken-listened) sequence, as shown in the linguistic representation of (2). A common denominator for the three representations can be given using the concepts of Predicate (P), or attraction relation, and Arguments (A), or objects of attraction, to express the stereo integration.

(2)	Stereo integration

Stereo integration cannot be broken into a binary integration (cup+bottle)+cup or cup+(bottle+cup). This ternary integration, in the classical linguistic sense, is neither binary, nor cyclic, nor derivational, nor hierarchical.

Based on this observation, the working hypothesis I will be following is that linguistic organisation might be using the same regulatory system which is utilised by sensory receptors to brake and to integrate bonds, generating percepts in doing so. If many specialised sensory receptors can integrate using the same regulatory system, then they may produce the same type of perceptions (Ramachandran 2001, Garcia-Bellido MS). These percepts may be made explicit in the environment by the motor system.

(3)

The research programme I advocate here is to find how, where and when sensory inputs are regulated producing internal percepts. The possibility that many factors work in parallel will be the assumption followed here (Jackendoff 2002, Seuren 2004). Language as a percept, may emerge under this approach, from the complexities of relations integrating, stabilizing and resolving, simultaneously and non simultaneously, in some perceptually coherent defined way. This complexity may produce many different parallel types of percepts. I will focus on an integrative system which regulates integration by achieving compression and decompression effects.

The structure of this chapter is as follows. I will first give concrete data and describe it according to some precise observations (5.2 and 5.3). I will then give some definitions to help situating the data and the observations within a linguistic-perceptual framework (5.4). I will suggest that a more complex integrative approach may be more successful to account for the effects found in many languages (5.5 and 5.6). I will then present the problems that current linguistic approaches generate when trying to account for variation in language (5.7). Finally, I will suggest that if the integrative similarities found for compressing segmental and word sequencing could be an homology not only within the linguistic system but also within a more general perceptual system which includes vision, then linguistic behaviour might just be a very old integrative perceptual system which has been motorly sequenced.

Índice general  I  Siguiente