Solving the borrowing problem when reconstructing grammar in a proto-language
The difficulty of re-constructing an extinct parent language, commonly referred to by linguists as a proto-language, without being trapped by the problem of grammatical borrowing, has been tackled by CoEDL postdoc Don Daniels in a paper titled "A method for mitigating the problem of borrowing in syntactic reconstruction" in Studies in Language.
In the paper, Daniels applies his method to two grammatical constructions from the Sogeram languages of Papua New Guinea.
Daniels writes that most scholars agree grammatical borrowing is a serious problem to grammatical reconstruction and proposes a step-by-step solution. This solution is to build up a proto-language grammar from constructions that contain similar phonological material, test that material for relatedness, and then to apply traditional linguistic analysis to those reconstructions.
“It is about reducing the odds that what we reconstruct is the result of contact-induced change rather than shared retention,” he says.
Daniels summarises the approach as:
- Establish correspondence sets for partially schematic grammatical constructions.
- Ensure that the phonological material reflects inheritance, not contact.
- Reconstruct the constructions.
- Where a set of similar constructions has been reconstructed, posit a generalisation over those constructions that accounts for variation.
His method builds on the approach that others have suggested (that grammatical reconstruction ought to rely on morphological reconstruction) and takes it a step further, by allowing for generalisations over sets of daughter constructions. This enables the reconstruction of more complete proto-grammar.
Image at top: Job Askai working with Don Daniels (right) on the Manat language in Madang, Papua New Guinea.