Erm, if at the 100% level "all sentences are just the word 'sustainable' repeated over and over", could these still be defined as 'sentences'?! Mark Liberman mentioned on this graph yesterday on the UPenn blog http://languagelog.ldc.upenn.edu/nll/?p=3723#more-3723 with some comparisons from Google Books.
Just a comment - what do we usefully learn about usage from graphs of single wordform frequencies? To me, a single wordform frequency-focused approach doesn't show much more than whether a wordform becomes more or less frequently used. A design engineer once asked me "how long have people been using green and sustainable with an environmental meaning?" and wasn't too impressed with the plain frequency results. The OED (historically crowd-sourced, currently corpus-based) told us that the emergence of the "environmentally sustainable" meaning of sustainable to around 1980, but I also found the search interface for COCA/COHA useful for providing more semantic information. e.g. once the notion that an activity is sustainable takes on a positive evaluation, different nouns come to be pre-modified by sustainable (not just development, agriculture, and growth, but business, production, tourism and policies). COCA also lets you explore syntactic shifts e.g. changes in how often sustainable is used attributively rather than predicatively.
Graphs of single wordform frequencies also ignore polarity: Jasienski (2006), for example, searched for "words indicating surprise" among 30 million abstracts of English-language scientific papers. He compared the frequencies of words appearing in these abstracts with their frequencies in the Brown corpus (!) and found that "the word 'surprising' appears 12 times more frequently in the natural sciences than in standard English," and concludes from this observation (p1112) that "the study of nature does indeed seem to surprise us. The odds of finding in abstracts of scientific research papers a result or conclusion described as 'surprising' ... are an order of magnitude greater than in standard language."
My examples below (from Electrical and Mechanical Engineering) also use the word surprising, and would thus be counted in Jasienski's results, but they're actually expressing a lack of surprise:
"The experimentally determined losses have relatively large error bars at the highest T, so it would not be surprising if the nominal result for int (300 K) is somewhat too high." (Electrical Engineering)
"Considering gravitational effects, the larger is d, the more efficient are gravity effects to deform the interface. It is thus not surprising that inertia l effects have to be increased to compensate for it." (Mechanical Engineering)
So if you make generalizations from the frequency of an adjective in a corpus then it would also be a good idea to check what's going on in its environment - polarity, pre-modification, complementation, collocation etc. To me, simply looking at frequency graphs for sustainable obscures the interesting stuff.
Jasienski, M. (2006). It's incredible how often we're surprised by findings. Nature, 440, 1112.
From: corpora-bounces at uib.no [mailto:corpora-bounces at uib.no] On Behalf Of Angus Grieve-Smith Sent: Monday, January 23, 2012 4:55 PM To: corpora at uib.no Subject: [Corpora-List] Fwd: increasing use of "sustainable"
Can you find something wrong with this picture?
-------- Original Message -------- Subject:
increasing use of "sustainable"
Mon, 23 Jan 2012 09:01:35 +0000
Dave Sayers <D.Sayers at SWANSEA.AC.UK><mailto:D.Sayers at SWANSEA.AC.UK>
VAR-L at JISCMAIL.AC.UK<mailto:VAR-L at JISCMAIL.AC.UK>
Happy Monday everyone.
-Angus B. Grieve-Smith
grvsmth at panix.com<mailto:grvsmth at panix.com>
David J. Oakey, BA, MEd, PhD
Applied Linguistics Program
Iowa State University
317 Ross Hall
Ames, Iowa 50011-1201, USA
Email: djoakey at iastate.edu
-------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/html Size: 10693 bytes Desc: not available URL: <https://mailman.uib.no/public/corpora/attachments/20120124/93d3c1e8/attachment.txt>