One of the most hotly contested aspects of modern syntactic theory is its emphasis on modular subsystems of grammar and, in the case of some authors, its postulation of a genetic basis for these subsystems. In one particularly popular theory, Government and Binding (cf. the Chomsky stack in your local library), linguistic competence is said to be due to the interaction of a number of modular principles which bear no relation to any other cognitive system, and which are ultimately due to genetic encoding.

Such an explanation, of course, has numerous problems, among which is the fact the reduction to genetics is, in effect, a machina ex deus argument: Phenomenon X is due to Subjacency,¹ Subjacency is due to genetics, and genetics is due to <?>, possibly some god, especially in Louisiana. In addition, of course, there is the fact that other human capabilities which have genetic causes appear to be comparatively minor alterations of faculties which exist in higher primates, so that the claim of species-specific syntactic uniqueness and genetic source are in conflict. The argument from genetics places the causation of language directly in the lap of the First Cause, or First Causes, and renders any further investigation impossible. This is not an a priori fatal condition for our field, however: we can embrace the situation along with all of its implications, and still build an interesting, enjoyable, and most importantly, irrelevant science of linguistics.

I would like to adumbrate (cf. Tyler, 1980-90 for the appropriate usage of this term) an interesting possibility. Humans do, in fact appear to have something that no other species, regardless of its cleverness, appears to have: mythology and religion. Let us, instead of positing machina ex deus arguments, return to the proud tradition of the deus ex machina. Such a move will have many major benefits. For example, gods are quite modular. Zeus never really agreed with Hades on anything, and they had nothing in common other than a distressing tendency to turn into animals and ravish maidens. Let us, for the moment, postulate the existence of Norbert, God of Phrasal Category Expansion, and Bloduwedda, Our Lady of Linear Precedence.² Now both Norbert and Bloduwedda are judgemental types, and will mercilessly stomp on any sentence which is not, in their view, Holy and Pure. They have, however, completely different notions of what constitutes kosherness in sentences: Norbert has given man commandments such as “Thou shalt not allow an adjunct to come between thy head and thy complement”³ while Bloduwedda has ordered that Articles Come First.⁴ Poor hapless humanity, however, has to make both Norbert and Bloduwedda happy, and so is severely constrained in terms of possible output. What is worse is that Bloduwedda is rather flighty, and makes different commandments for each language. In short, the entire pantheon of deities has to approve each sentence, but each deity operates according to different criteria, especially when drunk.

This brings us to the next point. A grammar based on deific influence inherently incorporates the competence/performance dichotomy. Deities, as is well known, have their random days. In canonical cases,⁵ Norbert will carefully ensure that no adjunct, however well-dressed or persuasive, will be able to warn his own or someone else’s head about an unmeant complement. Sometimes, however, Norbert will be overcome with ennui, and unable to take it anymore, will turn into a swan and ravish Leda.⁶ This is known as ‘performance’. Likewise, Vandrigon of the Many Roles, who oversees assignment of participant roles to all good little NPs, may occasionally suffer a fit of pique and punish a perfectly well-behaved NP with some heinous hardship role, such as Postdoc.⁷ Violations of proposed Divine Laws, etc., are thus easily explainable in terms of deific hormone surges. This explanation allows us to preserve the First rule of Linguistics, which is never let bad data interfere with a good theory.

A final, but quite important beneficial aspect of the proposed system is that it places linguistics consciously with the context within which it has always functioned. No one expects religion to have any basis in reality. Divine laws are divine laws because some famous powerful person, who may or may not be dead, says they are. If you disagree, you don’t get into the priesthood. The really important thing is: good ritual. People love complicated symbolism, and it is the job of the Hierophantic Linguist to give it to them. Phrase structures should always be governed by interesting restrictions, such as the Principle of Three Branches⁸ or the Maxim of the Intransigent INFL, which stipulates that INFL (a genius locii, somewhat like a hamadryad, which inhabits the higher peaks of phrase trees) must always be agreed with or he’ll make life hell for you. Note that such a system also allows for endless variation. Seemingly endless progress can be made by engaging in numerous schismatic alterations of Accepted Doctrine. Junior Faculty may, for example, develop a theory wherein Subjacency is seen as an evil plot by a Demiurge, and refuse to have any children or to eat beans until they get an office with a window. Binarists and the Servants of the Underspecified can engage in stimulating shouting matches at major conventions, and even liven up the proceedings by attempting to rearrange one another’s features. Rumors can be be circulated about mysterious Zero Elements, which are unseen but which exert profound influences on daily life.⁹ Of course, much of this dialogue should, when possible, be carried on via photocopies with extremely limited distribution, so as to maintain a gnostic hierarchy.

In summary, I suggest that a pantheon-based theory of grammar will not only account for linguistic phenomena more honestly than current theories, but also allow for more interesting methodologies in our field. The exact structure of such a grammar need not be specified here, as it follows from general principles, and if you don’t already know you aren’t ready to know.

². Although these deities are being introduced as if for the sake of argument, it should be stressed that their actual existence was revealed to the prophet Blimfozz (“The Maximally Bracketed”) in the holy city of Scranton.

³. Carefully translated, the original text of this commandment can be interpreted as “Never let your aides tell you that someone didn’t really mean it when they told you your budget plan was brilliant.” Subsequent application of this principle to syntax was, apparently, a result of scribal enthusiasm.

⁴. Also known as the ‘reading list’ principle.

⁵. ‘Canonical Case’ is a term used to refer to sentences whose structures are in accordance with linguistic competence, i.e., sentences which do not contradict the theory.

⁶. This is, of course, completely reprehensible behavior. The term ravish in mythology cannot, however, be identified with either sex or rape, since (A) maidens get ravished in Victorian literature, and, as is widely known, the Victorians did not have procreative organs or a sex drive, and (B) swans can ravish people. The suggested gloss is something like ‘to salivate on someone while making enthusiastic cooing noises’. Leda, apparently, became rather inured to this behavior, judging from the fact that she was ravished in a large number of Victorian poems. She is, incidentally, credited with the invention of the raincoat.

⁷. Vandrigon, like many of the Greek gods, is far more interested in whether an NP is well-formed than well-behaved.

⁸. This principle relates to the Three Types of Cow which were Not Stolen by Gwydion (He instead stole pigs, shoes, and weapons, respectively, and really off-put Arianhrod, who was busy drumming Subjacency into one of her Junior Faculty). He was drunk at the time.

⁹. There is, in fact, a form of divination known as Zeromancy which purports to chart a person’s character based on the alignment of Empty Categories at his birth.

	Advancing Pedagogical Methodology in the Linguistics Classroom--Shr Ji Wei
	Discourse Antipassivity in Indonesian--Mokele Mbembe
	Linguist of Fortune -- JLSSCNC Vol I, No 1 Contents