Crowdsourcing grammars, or the death of the (grammar) book. WAS: Publishing; Mouton "discounts" for ALT members

[Everett, Daniel]

It was interesting to read David's remarks, with which I fully agree, about the continued need for grammars and why it would be a mistake to think that they could be replaced by data bases. I had just heard from a psycholinguist the other day that linguists have no business writing grammars. That they should only provide data bases. The reason? Linguists are not trained in scientific methodology (i.e. psychology) and therefore their grammars are nothing more than speculations. Leave out the speculation and you have a data base. 

The problem with the reasoning, aside from the fact that grammars are much more than speculations, is that databases reflect point of view and theory as much as grammars. Unless they are nothing more than dumps of all the data one has ever collected. I suppose that we could take the garbage-dump + scavaging approach to language work. But for the reasons David mentions, and more, this is probably not a good idea.

Dan


On Nov 16, 2011, at 2:40 AM, David Gil wrote:
>> 
> I think many linguists are by now completely convinced that databases
> are a wonderful tool that can be used in many ways that would have been
> completely inconceivable just a few short decades ago.  (Personally
> speaking, I think it would be fair to say that FileMaker Pro is perhaps
> my favourite application these days.)
> 
> Nevertheless, I would disagree that databases can or should replace a
> narrative text such as a grammar book whose most appropriate
> representation is (to the best of my knowledge) still in formats such as
> Word and Pdf.  What we do, as linguists, is not just collecting and
> organizing linguistic data, but also constructing generalizations and
> theorizing on the basis of such data.  Spinning a story, if you will.
> This is how our minds work, this is how we think.  And the stories that
> we spin necessarily assume the form of a text, written in a language
> that we assume our prospective addressees will be able to read.  To
> which other linguists respond with other texts.
> 
> Now clearly, there will be, and in fact already are, new developments in
> how such texts are represented.  For example, moving from dead trees to
> silicon enables us to create and consume hypertext, in which the
> narrative is no longer embedded in a single linear progression.  (This
> is what footnotes do in a rudimentary fashion, but hypertext will do
> this much better.)   But as I see it, such hypertexts will still be more
> like today's books than today's (or tomorrow's) databases.
> 
> So to conclude, databases are wonderful, but we'll still continue to
> need books.  Almost certainly our books will evolve, both in material
> form and internal structure; but some descendant of today's books will
> continue to be around, and to be used, by linguists (and presumably
> other scientists) alongside our databases.
> 
> In particular, if I wish, at some point in the future, to get a quick
> (or not so quick) picture of some unfamiliar language, I would still
> like to have at my disposal not just a database to work with, but also
> an introduction to the language, in text form, produced by a linguist
> who has already invested time and energy into thinking about how the
> language works — in other words, a 21st century version of a grammar by
> Mouton or some other publisher.
> 
> David
>