[Corpora-List] Do you think LINGUISTICS is SCIENCE or ARTS?

Dominic Widdows widdows at google.com
Tue Mar 23 15:05:01 UTC 2010


On Tue, Mar 23, 2010 at 10:51 AM, pascale <pascale at ee.ust.hk> wrote:
> Dear colleagues,
>
> Whether a discipline is science or not depends on the definition of science.
> It would be fair to say that anything claims to be a science should be based
> on scientific methods. The definition of scientific methods is also pretty
> clear since
>
> According to Merriam-Webster:
>
> Main Entry: scientific method
> Function: noun
> Date: circa 1810
> : principles and procedures for the systematic pursuit of knowledge
> involving the recognition and formulation of a problem, the collection of
> data through observation and experiment, and the formulation and testing of
> hypotheses

But the main distinction here is that science is about knowing things
rather than making things, which is nowadays the difference between
science and technology. Part of what an architect or engineer must
know is whether a building or bridge will stand up, and for this
scientific knowledge is vital. But it's not the whole story of being
an architect or engineer.

I'd say the same about writing a piece of music. I know why a C and a
G "go together", and I don't think this knowledge is any less
scientific than knowing that the middle of an arch will bear weight
more effectively than the middle of a flat plank of wood.

> Whereas I get the beauty of music and the relationship between rhythm and
> melody and octaves, harmonics, etc. I would dare say that musicians do not
> compose music using the above-mentioned "scientific method". (Whereas signal
> processing engineers who analyze music and come up with scientific
> hypothesis and then data and then models to classify a piece of composition
> as either "music" or "junk" is at least doing science. I realize how
> self-serving for engineers this statement could be though.)

So in the sense above, the signal processor is a scientist because
their goal is descriptive knowledge. The same could be said of a
student of musical analysis. The composer, architect and programmer is
an artist / technologist because their goal is creative. They must use
scientific knowledge, but they must do more than this as well.

> As such, if we were doing linguistics using the "scientific method", then it
> is science. If not, it is not. This is why there are different approaches in
> linguistics.
>
> Having said there, there is nothing judgmental in labeling something as
> "science" or "art". Therefore, saying that XXX (e.g. mathematics,
> linguistics, music) is not science is simply following the standard
> definition of "science", not a criticism.
>
> Of course, one can dispute the definition of the term "scientific method"
> today. Then this would be a discussion at a totally different level, and
> that discussion itself would be art rather than science as it is defined
> today.
>

I agree completely. And please understand that I don't expect anything
I've written above to be accepted as "scientific truth"!

Best wishes,
Dominic

> Best regards,
>
> Pascale Fung
>
>
>
>
>
> Dominic Widdows wrote:
>>
>> On Tue, Mar 23, 2010 at 9:34 AM, Gill Philip
>> <g.philip.polidoro at gmail.com> wrote:
>>
>>>
>>> dear Corpus colleagues,
>>>
>>> Just while we're off-topic, remember that Cage's 4'33" *proves* that
>>> music is music even when there's no sound and no rhythm but only the
>>> act of performance.
>>>
>>
>> Alternatively, it proves that Cage is trying to make headlines rather
>> than music.
>>
>> The traditional quadrivium is arithmetic, geometry, music, and
>> astronomy - an interesting mix that predates our modern idea of
>> "science" as distinct from natural philosophy. Any mathematical
>> musician worth a grain of salt would have been able to tell Gauss that
>> some of the "nice" was because of the relationship between octaves,
>> fifths, harmonics, and before you know it you're going straight back
>> through Plato to Pythagoras.
>>
>>
>>>
>>> Anyway, Yuri's question is interesting.
>>> I suppose the anwer depends on what you consider linguistics and its
>>> object of study to be.
>>>
>>> If, like many on this list, you use data samples to study language and
>>> meaning empirically, then 'linguistics' is essentially a social
>>> science
>>> If you prefer to concentrate on language in your mind, then
>>> 'linguistics' is a form of philosophy or logic which makes it firmly
>>> rooted in the humanities
>>>
>>
>> But papers and books written in this rationalist tradition look like
>> more mathematical than Newton's and even Russell and Whitehead's
>> "Principia"! It's such a shame that scholasticism turned logic from a
>> very practical to a very theoretical subject, when Aristotle invented
>> logic I believe it was partly to support his practical reasoning about
>> cows and ants and shellfish - the notion that logic and empiricism
>> don't go together is a sad tribute to how impoverished the followers
>> can become compared with the founders.
>>
>> To my mind, all this demonstrates pretty quickly that "science or art"
>> is an anachronistic post-industrial question. I think a lot of it
>> comes from the belief that quantitative research is more likely to
>> lead to technological advancement, hence more likely to get funding
>> from governments. So if you want jobs you want to be a science, right?
>>
>> So the difference between calling something science and art is mainly
>> political.
>>
>> Best wishes,
>> Dominic
>>
>>
>>>
>>> And if you're using linguistic data but not doing linguistic analysis
>>> or contributing to the study of language, you're probably doing a form
>>> of science... but some might argue that you're not doing linguistics
>>> either
>>>
>>> Gill
>>>
>>>
>>> *********************************
>>> Dr. Gill Philip
>>> CILTA
>>> Università degli Studi di Bologna
>>> Piazza San Giovanni in Monte, 4
>>> 40124 Bologna
>>> Italy
>>>
>>> _______________________________________________
>>> Corpora mailing list
>>> Corpora at uib.no
>>> http://mailman.uib.no/listinfo/corpora
>>>
>>>
>>
>> _______________________________________________
>> Corpora mailing list
>> Corpora at uib.no
>> http://mailman.uib.no/listinfo/corpora
>>
>
>

_______________________________________________
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora



More information about the Corpora mailing list