A query...

[Daniel L. Everett]

Miriam,

You correctly attributed the things to me that I said.

We have to change the culture of the field if we are going to provide  
the documentation for replicability that we need. Lack of money is  
not an excuse. We need more - more money and more replicability.

In fact, there in Scotland you have access to research grants that  
far exceed the grants available to linguists from the NSF by and  
large. So the UK might actually set the standard on these things.

As to what files you can make available or should make available, of  
course there are limits. No one says that 100% of data should be  
there. This is the decision of the speech community in any case. The  
ideal for us as linguists is that there be enough data available for  
any of us with an interest to redo any study you have done, at least  
for languages in which access is very difficult and especially for  
endangered languages. For English this is not as crucial, perhaps.

I am quite concerned about linguistic typology and theory when they  
base their conclusions on grammars that we haven't got the data to  
replicate the analysis of. And that is what we generally do. Mainly  
because that is what we are forced to do. We 'take the word' of the  
grammar writer for their data because we have no record of it  
accessible to us.

How do we know that the old Jesuit grammars, often of extinct  
languages, are good? Partially by archiving.  In the case of  
Anchieta's grammar of Tupinamba, for example, we have additional  
data, the data in his own catechism and dictionary and the  
conversations 'recorded' by Jean de Lery, the French Calvinist. Ffor  
most of the others, we have the modern languages to compare their  
grammars against.

It is true that we would be better served if there were a standard.  
But if the archiving system is non-proprietary and if it has clear  
instructions, then I am very glad for its existence, even if it has  
problems and is non-standard.

There are problems and they won't be overcome in a year or two. More  
money for linguistics research may not be forthcoming. But I'd rather  
see bigger grants going to fewer research projects that lots of small  
grants that leave little hope for replicability of the results.

I just cannot see the problem here.

-- Dan

P.S. Some readers might find it instructive to compare grant sizes of  
the different funding agencies. In the UK linguistic awards of up to  
roughly one million dollars (AHRC) or 1.5 million (ESRC) for five  
years are allowed.  For the NSF linguistics awards are usually less  
than one hundred thousand dollars per year, for a preferred three- 
year maximum as I understand it. Things may have changed. More money  
is available from the NSF in principle since, last award I had at  
least, it had few explicit caps.

Here are three pages, for NSF, AHRC, and ESRC.

NSF: http://www.nsf.gov/div/index.jsp?org=BCS
(then go to award search and just type in linguistics and scroll down)

AHRC: http://www.ahrb.ac.uk/awards/

ESRC: http://www.esrc.ac.uk/ESRCInfoCentre/help/research_list.aspx#skip




**********************
Daniel L. Everett, Professor of Linguistics & Anthropology and Chair,
Department of Languages, Literatures, and Cultures
Campus Box 4300
Illinois State University
Normal, Illinois 61790-4300
OFFICE: 309-438-3604
FAX: 309-438-8038
Dept: http://www.llc.ilstu.edu/default.asp
Recursion: http://www.llc.ilstu.edu/rechul/
Personal: http://www.llc.ilstu.edu/dlevere/

and

Honorary Professor of Linguistics
University of Manchester
Manchester, UK