J ackendoff ’s (2010) caveat “Your theory of language evolution depends on
your theory of language” must be seriously reckoned with.
Highly relevant here is the question of whether protolanguage existed
before the advent of human language. In fact, there is suggestive evidence
for its existence from a variety of animal cognition experiments and also
from studies of human pathological conditions. Primates or non-primates,
several non-human species have been shown to have the capacity for non-
hierarchical, linear grammar. Dolphins, for example, can discriminate differ-
ent meanings of a pair of sentences depending on the order of the words
in them (“words” here are represented by gestures and electronic sounds);
they understand how take the ball to the hoop differs from take the hoop to
the ball (He rman 2010). This is, of course, not to say that dolphins have
Merge, since Merge-based syntax will yield a grammar that is dependent on
hierarchy and not on linearity.
In humans, Broca’s aphasics from time to time exhibit a similar behavior.
They correctly understand the dog chased the cat but the cat was chased by the
dog only at chance level. In the past, this situation was linked to the idea that
the patients had some trouble processing a structure generated by movement
(Y. Grodzinsky’s trace-deletion hypothesis being a concrete proposal), an idea
which is no longer an option. For one thing, given the VP-internal subject
hypothesis, deriving an active sentence involves movement, too. For another,
movement is now understood to be a case of Merge (internal Merge) and has
no special status in the minimalist theory of syntax.
The correct interpretation of this degraded performance by Broca’s aphasics,
then, seems to be this. Merge-based hierarchical structure processing ceases
working, and instead a simple processing strategy based on word order comes
to the fore, according to which the subject (agent) must precede the object
(theme or patient). This explains not only the active-passive contrast above, but
also why a relative clause like the cat that chased the dog (was small) can be cor-
rectly understood but the dog that the cat chased (was big) cannot, or even why
Chinese aphasics perform in the opposite way (because in Chinese the relative
clause precedes the head nominal) (see Grodz insky 2004).
Taken together, these obser vations point to the possibility that non-hierarchical
linear grammar is an evolutionary precursor of hierarchical grammar, and that
Figure 9.1 The minimalist design of human language, with an interface asymmetry.
Fallacies in evolutionary linguistics 143