[Lingtyp] On policing linguistic examples in Lingtyp Digest, Vol 90, Issue 27

[Adam Singerman]

I'd like to respond to two of Ian's comments in this very interesting
and thought-provoking thread.

1. "[H]itting someone is a bad thing to do. But I don’t think
linguists should avoid certain sentences because they depict something
morally bad."

All of us avoid certain sentences / certain verbs because of moral
objection. Most brutal acts are never represented in our linguistics
examples, and for good reason. So the question isn't *whether*
linguists should avoid certain sentences / verbs, as if doing so were
an all-or-nothing choice. The question is instead: where do we draw
the line between the acceptable and the unacceptable, given that we're
going to have to draw the line somewhere? It seems reasonable to err
on the side of choosing examples that do not portray speakers of
Indigenous, understudied, or under-documented languages as violent,
since violent stereotypes often get used to justify oppression and
discrimination.

2. "Is 'hug' and 'carry' as transitive as 'hit', though? According to
Hopper and Thompson (1980), one of the factors of transitivity is the
affectedness of the patient. Hugging or carrying someone does not
affect them as much as hitting does."

It's true that 'hit' may behave more transitively than 'carry' or
'hug', but that will not be true for all the relevant structural
properties. For any individual language you are likely to be able to
identify grammatical and semantic phenomena that operate according to
a more categorical intransitive-transitive distinction. If a given
grammatical property doesn't differentiate between a violent
transitive verb, like 'kill,' and a not-at-all violent verb, like
'feed,' why not illustrate that property using the less violent
option? And if a given grammatical property *does* differentiate
between 'kill' and 'feed,' that fact ought to be stated and justified
with examples.

When we linguists share our findings in publications and
presentations, we can only choose a subset of the data we have on hand
(sometimes a very, very, very small subset of the data we have on
hand) to illustrate a given point. I don't see any harm in asking
ourselves to be mindful of how our choice of data can be interpreted.

Best,
Adam