[Lingtyp] Conscious choice of linguistic examples
oesten at ling.su.se
Tue Mar 22 09:56:00 UTC 2022
Just a reminder that proper names have a rather low frequency in spoken discourse, at least in European languages. In a corpus of spoken Swedish representing about five hours of conversation, I did not find a single example of a transitive sentence with proper names in both subject and object position, that is, the classical "John loves Mary" pattern (Dahl 2000). In natural speech, the overwhelming majority of transitive subjects are pronominal and the overwhelming majority of transitive objects are inanimate. So a better strategy than worrying about what proper names to use may be not to use them at all.
Dahl, Östen. “Egophoricity in Discourse and Syntax.” FUNCTIONS OF LANGUAGE, vol. 7, no. 1, Jan. 2000, pp. 37–78.
Från: Lingtyp <lingtyp-bounces at listserv.linguistlist.org> För Sebastian Nordhoff
Skickat: den 22 mars 2022 10:36
Till: lingtyp at listserv.linguistlist.org
Ämne: Re: [Lingtyp] Conscious choice of linguistic examples
On 3/22/22 09:18, Maia Ponsonnet wrote:
> But this precisely assumes that "John" and "Mary" are unmarked first names.
Yes, that's the assumption, and that assumption is empirically warranted. If you have read the last 50 years of linguistic literature, your assumption, all else being equal, will be that in a new example with one referent, that referent will be "John", and if there are two, they will be "John" and "Mary".
If you have read the last 50 years of cryptographic literature, it will be Alice, Bob, and Eve.
To be clear, I am not advocating that this is a particularly desirable state of affairs. I just want to point out that "minimise astonishment/variation" is a principle which can guide didactic choices and has done so in the past. There are of course other, competing, principles, which one could rank higher.
> Which I do not believe to be accurate anymore in the context of a
> student population in Australia for instance (speaking of what I know).
> Presuming unmarkedness implicitly postulates a certain category of
> population as a norm (including color, age, probably social class...).
> And it seems even more problematic to consider that "John hits Mary"
> is unmarked 😉.
> In my opinion, it is worth risking to distract readers/students ever
> so slightly to avoid perpetuating stereotypes.
> Plus distraction can be a good thing!
> I have found that when presenting tedious grammatical phenomena about
> a minority language, semantically rich examples can be used as our
> "window" onto speakers' personalities, lives, cultures, inclinations
> etc. It is also a way to make speakers visible as individuals, and
> show respect, as pointed out by Felicity.
> In teaching contexts, such rich examples often trigger background
> questions that allow to give flesh to somewhat disincarnated theories.
> This sort of distraction is a good pedagogical tool to maintain
> interest and attention. As a reader, I find this applies to the
> scientific literature as well.
I personally find "John gives Mary the book" sentences incredibly dull, but I could see why for certain types of highly formalised intricate syntactic argumentation you really want to minimise variation.
If I were to write intricate syntactic papers, I would probably go for Agatha for Agents and Pat for Patients. Maybe throw in some other names with A... and P... (and T... and R... for other roles). To offset the historical imbalance, I would personally not use John or any male names for this kind of paper. John has done a lot of work in the last 50 years, he can rest a little bit and leave the floor to others.
> Cheers and kind regards, Maïa
> Dr Maïa Ponsonnet
> Adjunct Researcher, Discipline of Linguistics
> Building M257
> The University of Western Australia
> 35 Stirling Hwy, Perth, WA (6009), Australia
> *From:* Lingtyp <lingtyp-bounces at listserv.linguistlist.org> on behalf
> of Sebastian Nordhoff <sebastian.nordhoff at glottotopia.de>
> *Sent:* Tuesday, 22 March 2022 8:17 AM
> *To:* lingtyp at listserv.linguistlist.org
> <lingtyp at listserv.linguistlist.org>
> *Subject:* [Lingtyp] Conscious choice of linguistic examples Dear all,
> consider the following illustrations of an adversative construction
> (1) My hat is red but my shirt is blue.
> (2) My hat is ochre but my shirt is teal.
> If your point is simply to illustrate the use of "but" in English, (1)
> is preferable to (2). Readers are not distracted by the use of
> uncommon colour terms, which have no impact whatsoever on the
> construction under discussion.
> When discussing grammatical phenomena, a common device to avoid
> distraction is to not vary participants in the article/book. This is
> why you get "John" and "Mary" all over the place. Upon seeing a
> sentence with "John" and "Mary", readers know immediately that the
> linguistic phenomenon to be discussed will focus on the words of the
> sentence which are not "John" and "Mary". This makes processing on the
> reader's side easier as compared to examples with a wide variety of names.
> This strategy is used in other disciplines as well. In cryptography,
> it is always Alice who wants to send a message to Bob, and Eve tries
> to intercept it. It would confuse readers if all of a sudden a
> different set of characters emerged and people would have to backtrack
> whether Marie-Pierre was the sender or the receiver.
> There is thus some didactic value in having
> The same didactic value holds for non-names, eg verbs or adjectives.
> "See" is the prototypical perception verb. One could use "smell", but
> readers should be faster in getting the stimulus-experiencer frame for
> "see". If your point is to talk about experiencers, start the
> discussion with "see" or "hear", not with "smell".
> To come back to transitive verbs, "hit" and "kill" are pretty much the
> "John and Mary" of the verbal domain. Readers will know that these
> verbs stand in for "affected patient" and "animate patient" and can
> extrapolate from there. You could of course use "tickle", but it will
> take longer for readers to process that the point you are making is
> "affected patient" and "animate patient".
> This is the didactic motivation.
> On the other hand, there are motivations of naturalness. Not all
> linguists think that the study of made-up examples of the type "John
> gives Mary the book" is a worthwhile exercise.
> Then, you get motivations of diversity and representation, which
> conflict with a fixed set of characters with traditional Western names
> and their roles. The characters' roles are furthermore stereotyped
> (that is the idea for Alice, Bob, and Eve to begin with), but of
> course breaking (or not perpetuating) stereotypes is also a motivation
> writers can have.
> So you get the competing motivations of didactics, naturalness, and
> Depending on the type of paper you write, one or the other of these
> motivations will prevail. But it is clear that this is a choice of the
> author whether "John hits Mary" or "Fatima tickled Li" are more
> suitable in the argument to be made. In line what has been said in
> this thread before, authors should realise that the choice of
> participants and verbs is in their power. This is not so much about
> policing (come on, no one will fine you!), but about realising which
> motivations can have an impact on your examples.
> Best wishes
> Lingtyp mailing list
> Lingtyp at listserv.linguistlist.org
Lingtyp mailing list
Lingtyp at listserv.linguistlist.org
More information about the Lingtyp