Linguistics as science and as academic discipline

[Spike Gildea]

This discussion is really interesting for me, both as a field worker and 
as someone who has been a department head, working with junior 
colleagues to present a strong case for tenure.  I'd like to take on 
Mark's question, and try to speak *just* to the issues that confront the 
linguist in the domain of the academic world (trying to set aside, for 
the moment, the ethical and human issues that are more specific to 
engaging with speech communities).  Both field methods (i.e., doing good 
science) and the tenure and promotion process (success in an academic 
job) are evolving constantly, both attempting to respond to changes in 
the world of research (many brought on by rapid advances in 
technology).  I appreciate the desire to leap ahead to a better new 
world, and to lambaste the stodgy old "system" that holds us back, but 
there are limitations to how fast change happens in the real world, and 
there are limitations to how much individuals can do -- as Claire says, 
we have to make choices, both those driven by our professional context 
(including the needs of tenure review committees) and those driven by 
our individual needs (the desire to publish on one aspect of a language 
takes time away from further work on another aspect).

With regard to field work, I think it is indisputable that linguistics 
has relied overmuch on analyzing black marks (sometimes in multiple 
colors in my data books) on a page.  These black marks are a partial 
representation of speech, which allow errors of representation 
(mistranscriptions and outright fabrications) to be introduced without 
providing a means for external reviewers to find and correct them short 
of going to the field themselves (prohibitively expensive, in both money 
and time  -- we don't have near enough linguists to do the job once, 
much less to do it twice!).  Further, in some field methods traditions, 
a data sentence can represent nothing more than a speaker responding 
"yes" to a linguist asking if the utterance could be said, but the black 
marks do not distinguish between such "data" (which in my tradition 
would not count as data) and actual utterances made in conversations 
between native speakers (which I hope everyone would agree counts as 
data).  In the last few years, technology has leaped ahead so quickly 
that it is now economically feasible to carry digital recording 
equipment and a laptop (and solar panels, where necessary) to the field, 
so that it is now possible to provide sound files for every utterance 
collected in field work.  Morey (2005) [Morey, Stephen.  2005.  The Tai 
languages of Assam - a Grammar and texts, Canberra: Pacific Linguistics, 
Research School of Pacific and Asian Studies, The Australian National 
University] provides an excellent example of a grammar that improves on 
the "black marks" method: he provides an accompanying CD where every 
example in the grammar and every line of every text is linked to a sound 
file, a recording of a native speaker saying the utterance.  Anyone with 
the CD can check all of these examples to verify any empirical detail 
that is relevant for a given analysis.  This sort of work provides a 
standard that all fieldworkers should aspire to, regardless of their 
position on web publishing.

But I am far from being able to achieve this standard in my own work and 
I have spoken about it with enough colleagues to know that, especially 
for those of us who were trained before this technology was available, 
even learning to use it at all is not a trivial issue.   In fact, not 
many people have done it, most that I know have developed idiosyncratic 
methods that work for them, but that may not be more generally 
applicable, and there are few (if any) venues where we can get training 
from someone who has a tested, generally applicable method.  As more 
individuals in the field try to engage with this new technology, there 
will be fewer unknowns, but while people with a greater affinity for 
technology jump in and experiment and debate about formats and methods, 
those of us who just want to be end users, who just want good tools and 
a straightforward model to follow, are concerned about entering into a 
massive project with no assurance that five years from now, we won't 
look back and see five years essentially wasted putting data into a 
format that didn't become generally accepted. 

In addition, field workers already require years to organize the black 
marks on paper (or in a computerized database) into grammars -- there is 
no question that adding the digital sound layer to this work will 
require a lot more time.  It is already a lot just to cut out sound 
files of examples, then link these sound files to written examples in 
some format (a toolbox database, Childes, even a simple Word document).  
Putting up our raw notes is unthinkable: we all have our own standards 
about how much of our "mess" we are willing to show to friends or 
strangers, but nobody will be able to just put up their database online, 
to make it accessible to others to read, without a lot of polishing 
work.  I agree, making recorded data available is the right thing to do, 
and I hope (and expect) that it's the future in linguistics, but right 
now, it's not established territory, and it's sure not a trivial 
decision for any individual field worker to start doing it.  There's 
only so many hours in the day, and most of us already have more on our 
plates than we can do justice to (more on this in a moment), and if we 
decide to dedicate a chunk of time to learning and then implementing 
this new technology, we'll have to make some hard choices about what to 
give up.  And ultimately, there are limits to how many skills one 
individual can master -- as Dan points out, to have enough work-hours 
and enough specialized knowledge to do this new, more extensive job, 
future models for field linguistics will almost certainly require teams 
of specialists to work together.

The other question is how academia rewards our labors.  First, most 
departments of linguistics do not have positions for those of us who 
view field work as our specialty -- most of us get hired for our 
theoretical specialty, for the classes in syntax or phonology (or 
whatever) that we can teach, in addition to the occasional field methods 
class. Once you've landed the tenure-line job, the tenure system is 
heavily stacked against fieldworkers even doing the more limited 
"traditional" kind of fieldwork.  The behaviors that are rewarded all 
involve refereed publications on paper (a few electronic journals are 
underway, but they are greatly outnumbered by print journals and, like 
any new journals, their value for tenure is still uncertain), and field 
work already takes substantial time away from that behavior.  Further, 
we must engage in disproportionate publishing activity in the specialty 
that we are hired for.  WIth that as the goal, field data collection 
provides a rather poor return for time spent.  While much of the time we 
spend collecting data and processing data is important to our 
understanding of the language, most of the data produced via this 
investment of time will not be important enough to show up even in a 
grammar; further, much of the data that will merit inclusion in a 
grammar will not prove relevant for current theoretical/typological 
debates, and hence the process of collecting and analyzing them is not 
rewarded until the grammar is published, which may not be prior to the 
deadline for consideration for tenure.  Time spent building extensive 
searchable databases of annotated material pays off down the road by 
making it easier to access the examples that are relevant for 
publishable articles, but the work of creating these databases doesn't 
count at all until the publications follow.  Until the system recognizes 
the validity of the contribution made by electronically accessible 
archives of organized data, untenured faculty really are not wise to put 
much time into cleaning up a database for web publication, and even 
tenured faculty might hesitate unless they are willing to set aside some 
professional ambitions. 

So how can we effect some change, help the system evolve so as to make 
it easier (and more rewarding) to do fieldwork in the more reliable ways 
that current technology makes possible.  To start with, I'd like to see 
a push for an academic culture that acknowledges the value (even the 
necessity) of CDs with sound files to accompany printed language data -- 
under that standard, I might not be able to publish anything for a few 
years, but I believe the reliability of the database available to 
typologists and theoreticians would increase sharply.  We also need to 
start a refereeing process for databases that are made available on the 
web -- they are as accessible as traditional publications, but they 
don't count for the tenure system because there is no peer-review-based 
gate-keeping mechanism to regulate the quality of the works that are 
disseminated.  This problem is not unique to field linguistics, but it 
might be a bigger issue for field workers to the extent that we are 
trying to create a new genre of publication, one that would not be 
possible offline.  Something like a reviews section in a journal, that 
lists and evaluates such sites, might provide the first steps towards 
gaining academic acknowledgment for such work.  On the side, we also 
really need to think about how to engage with non-fieldworker academics 
to convince them that what we do is scientifically valuable, and to 
educate them about how our activities differ from those of some other 
academics (including those who practice the currently dominant models of 
linguistics).  Recently, popular publicity about language endangerment 
has gone some distance towards justifying our work, but fieldworkers 
(and functionalists) have been out of power politically in linguistics 
departments for quite some time, which means our needs have not been 
made a priority inside most departments, and also that we have had 
little opportunity to shape the perception of the field on the part of 
academic administrators.  To implement changes in the system that will 
meet our needs, we will need the support not just of university 
administrators, but also of the less fieldwork-oriented colleagues in 
our own departments (and in departments that have not yet opened their 
doors to fieldworkers).  Put more bluntly, we need to figure out how to 
get theoretical linguists to (a) decide to replace a retiring colleague, 
say a phonologist or a syntactician, or a psycholinguist, with a 
fieldworker, (b) agree as a department to establish a separate standard 
for reviewing the academic output of fieldworkers, and (c) argue to the 
Dean, the Provost, and the various Promotion and Tenure committees that 
these standards are justifiable. 

Whew!  Thanks for a stimulating and diverting discussion!  I'm going 
back to work some more on my black marks on a screen...

Spike