Part of what makes (1c) and (5c) interesting is how characterization serves as an invitation to play. The speaker is practically obliging their interlocutor to participate in the ascription of interiority (to the shirt, to the bottle of rice vinegar). Consider again
5. (a) Where do you live?
(b) That pardalotte out back—where do they live?
(c) Where does the rice vinegar live?
In (1) we observed predicates of action. In (5), by contrast, we observe predicates of experiencing—specifically, the experience of dwelling.
(5c) is distinguished by a mismatch between our semantic template for experiencers of dwelling and the semantic profile of a bottle of rice vinegar. But we would hardly call this expression awkward, unless, perhaps, the speaker were new to our shared language and we suspected they lacked the resources to formulate the question in a more conventional way, e.g.,
11. Where do you keep the rice vinegar?
If anything, (5c) feels more fluent by virtue of its playfulness. The formulation both acknowledges and affirms the speaker’s commitment to a potentially uncomfortable activity (shared meal prep in a setting familiar to the hearer but not to the speaker etc).
This is a not uncommon form of verbal play. As speakers we avail ourselves of a range of strategies to signal when we’re being playful and when we’re simply describing events that diverge, in some fashion, from our expectations about the patients, agents, and experiencers of different kinds of events—“what can do what to what” and “what can experience what”. Here are two such strategies:
12. (a) I stroked the thylacine between the ears.
(b) The ferns brushed against us as we moved past.
(c) I was hit in the head by a tree limb.
(12) exhibits three common constructions for predicates of contact. In (12a), the default reading—to my ear and, I’d warrant, most speakers’—is that I intended to stroke the thylacine between the ears.
(12b–c) involve valence-lowering operations—transforming a transitive or two-place predicate into a one-place predicate. But these are valence-lowering operations of two distinct kinds. In (12b), the agent, the ferns, remains in the external (“subject”) slot and the patient, us, has been shifted into a complement, against us. Here the default reading is that the contact was not so much unintentional as lacking in the possibility of intention—ferns are not the sort of thing we think of as capable of lying in wait—but in any case incidental: the contact unfolded in the course of another process—*as we moved past*—and had no material effect on that process. The fact of the contact’s lacking the potential for intention is indexed (pointed to), in (12b), by the shifting of the patient into a contressive (toward and into contact with) complement—*against us*.
In schematic terms, by choosing (13b) over (13a)—
13. (a) A brush P
(b) A brush against P
—we’re flagging the agent as somehow atypical of agents of contact, which we tend to treat as possessed of intentionality. Consider
14. (a) I brushed the thylacine with a stiff-bristled brush.
(b) The ferns brushed us as we moved past.
In (14b), by slotting the ferns as the agent of a transitive brush construction, we’re evoking a semantic profile for the ferns closer to the semantic template for the predicate brushed. The change is subtle—for most listeners, (14b) would not evoke an attribution to the ferns of a degree of intentionality comparable to that evoked for I in (14a). But here—to my ear at least—*the ferns* do seem to be intending something in a way that is not the case in (12b).
In (12c) we’ve adopted a complementary strategy of valence lowering. Here the agent—*a tree limb*—has been demoted from the external slot to an instrumental or agentive complement, by a tree limb. The verb, meanwhile, has shifted to a passive form, with the patient, I, in the external slot. In schematic terms,
15. (a) A hit P
(b) P be hit by A
By demoting the agent and promoting the patient, we’re signaling that the event, while not incidental to the patient’s subsequent experience—in promoting the patient to the external slot we have, in fact, topicalized it, that is, signaled to listeners that it is now the topic of running discourse—was not a product of volition on the part of the agent. Contrast
16. The tree limb hit me in the head.
which exhibits the same reharmonization of agent profile and predicate template as (14b), evoking a shadow-intentionality in the tree limb where it would ordinarily be lacking.