Since Chomsky’s (1981) revision of the Projection Principle (PP) to include the stipulation that every sentence have a subject (Extended Projection Principle; EPP), syntactic theory has been troubled with an apparent explanatory gap. The empirical coverage of the content of EPP is well-documented (Carnie, 1995; Chomsky, 1994), but it remains restricted to a descriptive statement. Explanatory accounts of why Universal Grammar (UG) should contain such a stipulation remains mysterious. This issue is compounded further in light of the minimalist program which sets forth a strict requirement in terms of computational economy – perform an operation only if you must, and then do so in the most optimal/cost-effective way possible. Philosophical and mathematical inquiry into the logical structure of linguistic theory have suggested that such stipulations as EPP are results of the general natural principles that govern all complex systems. One far-reaching claim, in this regards, concerns Uriagereka’s (2000) argument that linguistic structures display the properties of the Golden Ratio, as seen in the Fibonacci numbers. Several studies since have suggested that when tree structures are maximized in terms of specifiers and complements, it results in a Fibonacci growth (Carnie et al. 2005; Medeiros & Piatelli-Palmarini, 2016; Medeiros, 2012), and also that realized as auditory sequence Fibonacci-ness of grammars are capable of eliciting special behavioral responses (Krivochen & Saddy, 2018).
Related, of course, is the notion of EPP itself. To briefly recap the line of argument Medeiros and Piatelli-Palmarini are pushing, it is claimed that the x-bar schema serves two essentially minimalist purposes — (i) it results in bushier trees that reduce c-command possibilities which leads to a reduction in search operations, and (ii) the schema boasts the Golden Ratio as its dominant eiganvalue which results in special behavioral preferences. The x-bar is, thus, assumed to provide a natural law based explanation for EPP, because the EPP-type projection helps satisfy the Fibonacci-ness of the template. All the subsequent discussion of the significance of the schema are dependent on EPP being core to human syntax, which if invalidated would significantly weaken our basic hypothesis itself. In the literature it has been sometimes argued that certain VSO languages do not manifest EPP (McCloskey, 1996). Furthermore, VSO languages are themselves known to be a typological rarity ranging between 9% (Mallison & Blake, 1981) to 9.2% (Tomlin, 1986) of attested languages. As such, investigation of the structural schema underlying phrasal projections in such languages presents an even more interesting case since, assuming a lack of EPP, these languages themselves could be argued to be exceptions to a general/universal naturally motivated phrasal fractal patterns. Not unlike the Fibonacci-ness of the spectral patterning of Zebra stripes exhibiting exceptions to the norm due to phenomena like the growth of the ears, or localized epidermal irregularities at birth. Indeed, in a rather technically interesting squib David Adger (QMUL, U.K.) discusses the EPP problem raised by McCloskey and colleagues, and suggests a strongly Chomskyan solution. The original squib can be found here: A labelling solution to a curious EPP effect.
While a little technical, and definitely not a paperback-read, the squib is highly insightful and revealing, so I highly recommend reading the original text for yourself (then following it up, may be, with Juan Uriagereka’s book and the papers by Medeiros, Carnie and Piatelli-Palmarini, should the curiosity bug bite!). But briefly, Adger revisits a problem McCloskey (1996) noticed: in certain (Irish) verbs single argument appeared either as a prepositional phrase or as a nominal phrase. In the former case, the PP appears low in the structure, while the nominal phrase moves to the standard subject position. This problematic to capture in a theory that assumes that movements must be triggered. Why? Because if the theory uses something to trigger the movement of the nominal subject, what happens to that trigger when the subject is a PP? Proposing something like “If it’s a PP then there’s no trigger”, is obviously devoid of any explanatory prowess. If movement is always triggered, then movement implies trigger, and in turn the presence of a trigger will, in principle, move a move-able constituent (by triggering movement!). The haiku-esque nature of the conundrum aside, Adger argues that some ideas from Chomsky’s recent POP framework is, indeed, capable of providing a rather elegant solution to the problem at hand. Overall, it suggests that there are no EPP features of heads, but rather that any phrases that can’t stay in situ would eventually end up wherever they can be licensed.