For some time now in American English, there has been a strong tendency to replace the traditional complement––here a postposition––after the verb head, namely for, by the formerly non-normative to, resulting in a contemporary vacillation between the two constructions. This variation can be used to illustrate many aspects of the entire process of change, each of which merits separate treatment.
Leaving aside all but the raison d’être, we can first compare the meanings of the alternating postpositions. To generally means ‘in the direction of’ something after verbs of motion, as in the prototypical go to, without, however, precluding the attainment of the goal of the verb. Hence, in a command to a child such as go to your mother, both the directional and the telic meanings are present: going involves both starting out in some direction and having a possible goal, although reaching that goal may not be explicit (as it is in go to school, to work, etc.). For after verbs of motion reverses the hierarchical relationship between directionality and telicity: in a colloquial American expression such as go for it, the directional aspect of for is completely subordinated by the meaning of attaining a goal.
In head for the telic aspect also predominates: the phrase not only means setting out in a certain direction (which is presupposed) but makes the attainment of a goal its primary content. It is in this respect that the phrase differs from head toward: the meaning of toward is essentially the same as that of to when they function as verbal complements. The difference is rendered explicit when the object is a quintessentially directional substantive, as with a compass point. Hence the contrast between head for the hills and head toward the east, with *head for the east being unidiomatic (cf. the complement-less head east).
From the point of view of linguistic structure, one might infer from the foregoing analysis that there is something about the semantics of the two postpositions that is at stake, specifically a difference of rank in the semantic syntagms associated with each of them. An analysis that trades in competing semantic hierarchies may not seem to constitute an explanation of the change from one syntactic pattern to another, but this is not strictly so. The nature of grammar is such that what appears in speech or is expressed can always be traced to underlying grammatical relations, which are semantic in their essence, as its cause.
But in the syntactic change discussed above, one unsatisfied with this type of intrinsic explanation might wish to speculate about causes inherent in the larger communicative situation. Although hard evidence is unavailable, perhaps the change has its transcendent explanation in the larger tendency within contemporary American culture to neutralize social hierarchies, i.e., to scant hypotaxis in favor of parataxis. With the encompassing social structure and its flux as a reference point, the change in grammar would find its place as a piece of worldmaking.
This kind of explanation may be extrinsic to grammar proper, but there is also no gainsaying that syntax connects with reality in just this sort of way. In the change at hand, to repeat, we would seem to have a leveling of hierarchy––parataxis triumphing over hypotaxis. The constituents being leveled in meaning are the two complements in conjunction with the two derived (figurative) meanings of head. Head participates with for in creating a compound meaning involving telicity by imparting its meaning/function as the locus of cogitation (thought, planning) to the phrase. When head combines with to, however, its contribution to the compound is limited to the aspect of directionality: setting out in a certain direction means having the “head” go first.