On 2013 Jan 29, at 8:41 PM, William Donzelli wrote:
Ngram is the
determining authority?
Not perfect, but it does the job. Suggest something better.
Even if there isn't anything better that doesn't validate Ngram for
the task.
Ngram is neat
but it may have it's own skew to results. A new or
'in-vogue'
term like midicomputer may show up more in trade rags than
published books
in comparison to more established terms like minicomputer and
mainframe.
Yes, that is true, but certainly there will be some migration over the
years from the rags to the books, just as it happens today. Jargon
works its way into the general language, even in something as awful as
a Tom Clancey novel.
It may do so, but it may also go from 'in-vogue' to simply out.
I remember "midicomputer" from the 70's, even though I was just a kid
at the time. My own perception or recollection, as I mentioned
earlier, was the term got dropped or overridden in favour of
supermini by the 80's.
I Ngrammed
'midicomputer' and 'bubblehead' (a 'popular' term that
I figure
wouldn't show up so much in published books) and they come up with
comparable percentages (at different times).
Try "apples" and "oranges".
In the context of my point, 'popular' vs 'more formal' terms and how
that might affect Ngram results, I did.
If a narrow,industry-specific term like 'midicomputer' can show
similar percentages to a generic, popular bit of jargon like
'bubblehead', it says something about its prevalence, or else about
Ngram.