The Cabecar language (Chibchan, Costa Rica) distinguishes a copula from an existential verboid. The copula dä ~ rä is used if the predicate is substantival () or adjectival (), including nominalizations.
. | Jé | jayí | rä | jó. |
Cabecar | d.med | man | cop | undertaker |
That man is an undertaker. |
. | Jé | duláklë́i | kjä́bit-ä́ | rä | yakéi. |
Cabecar | [d.med | boy | flirt-ipfv] | cop | bad |
That boy’s flirting is bad. |
Nominalization of a clause, be it oriented or non-oriented, involves no grammatical apparatus whatsoever. The clause just takes the position of a noun phrase in the matrix clause. In , it is the [bracketed] subject of a copula clause.
The construction of may be interpreted in the same way as , with the difference that the subordinate clause of is (semantically) oriented towards a purpose adverbial depending on the verb ‘need’, as brought out by the literal translation.
. | Kalwá | kia-r=ké̱ | rä | dalí | ts-a̱-klä. |
Cabecar | [horse | want-mid(ipfv)=ipfv2] | cop | [load | transport-vsn-fin] |
The horse is needed to carry loads. | |||||
Lit.: | What the horse is needed for is to carry loads. |
In this literal interpretation, is, thus, a pseudo-cleft construction. However, since overt marks of subordination or relativization are lacking, the construction may be reinterpreted as a mono-clausal topic-comment structure where the erstwhile copula follows the topic, delimiting it against the comment and serving consequently as a thematic structure articulator in the configuration [ topic – TSA – comment ].
example | Kalwá kia-r=ké̱ | rä | dalí ts-a̱-klä. |
input | [ oriented open clause ]S | COP | [ focal component ] ]S |
output | [ predicate phrase | TSA | purpose adjunct ]S |
As usual, the reanalysis carried out becomes apparent when its original contextual conditions are no longer given. In , the erstwhile copula follows the actor noun phrase of a transitive clause.
. | Jímá̱k | te | rä | óshkoro | jé̠r | jé=bä | ñ-é̠. |
Cabecar | weasel | erg | tsa | chicken | liver | d.med=lim | eat2-ipfv |
The long-tailed weasel only eats chickens’ livers. |
This cannot be a copula clause. First, the subject of the putative copula clause involves no nominalization, so it cannot be a pseudo-cleft construction. Second, while the predicate of could, in principle, be nominalized, it could still not be the predicate of a copula clause because the actor NP of is marked as ergative, which fits the actor of the final transitive verb, but not the subject of a copula clause. So here the erstwhile copula is just a thematic structure articulator (tsa), a formative which separates the thematic portion of a clause from its comment portion.
Synchronically, the situation is as follows: In a clause containing the formative dä ~ rä, either the topic or the comment may be sentential and may contain the verb. However, if the topic contains the verb (), the construction is indistinguishible from a pseudo-cleft sentence, so the formative may be the copula. If the comment contains the verb (), the formative is just a TSA.