Discussion Groups / English Usage / January 2010

I'm not sure what you regard as easy, but one technique that works
quite well is to suggest in this group that some usage is very modern
and someone -- typically Evan Kirshenbaum or Donna Richoux, but maybe
someone else -- will tell you that it dates to 1731. Of course, you
have to choose examples that such people consider interesting.

This is called the "Recency Illusion".  When you notice a change, it
(almost tautologically) is perceived as being recent, because you've
only just noticed it.  In most (but not all) cases, the only thing
recent is your sensitivity to the usage in question.

-GAWollman

It may be impossible to ascertain when a usage started in speech. At
best we can say when it was first recorded in writing. There is
technology that allows us to search for this information. One way is to
consult the Oxford English Dictionary. That will tell you the earliest
example the editors have found of a usage. For example, the meaning of
"get" in the context you describe is defined thus:

"34 b. With passive pple.: To cause or procure oneself to be treated in a
certain way or to undergo a certain action; also, in weaker sense, to
come to be the object of a certain action. Often taking the place of be
as a passive-forming auxiliary where a continuous state is not intended
to be expressed."

This is followed by examples of the construction:
1652 "got acquainted"
1793 "got moored"
1814 "got supplied"
1823 "got entangled"
1826 "got shelved"
1848 "gets used to it"
1867 "got confounded"
1881 "get married"

If you search Google Books you might find earlier examples than those in
the OED, but you have to be careful; the dates stated by Google are
sometimes wrong.

As for "need" replacing "should," "ought to," or "has/have to", that is
covered by the OED definition:

"IV. 10. To be under a necessity or obligation (esp. one felt to be
self-imposed) to do or to be something."

The earliest examples of this are from the 14th century.

That one's interesting, according to the OED.  They cite it back to
before 1398:

   a1398 J. TREVISA tr. Bartholomaeus Anglicus _De Proprietatibus
   Rerum_ (BL Add.) f. 102v, A good phisician nediþ to loke wel
   a-boute and be ful ware.

but they say that it bacme rare between the mid 17th century and the
early 19th century.  Their quotes from the latter era start with

   1834 H. MARTINEAU Moral II. 50 The capitalists do not need to
        combine when labour superabounds.  

Things like

   Capt. Fordyce, of the grenadiers, led the van with his company,
   and Lieut. Batut commanded the advanced party.  The former got
   killed within a few yards of the breast-work, with 12 privates.
   The lieutenant, with 16 soldiers, were taken prisoners, all
   wounded.

                     _The Universal Magazine_, February, 1776

?  

Actually, it's far easier now than it used to be.  The first thing to
do is to see if you can get on-line access to the OED.  This will
typically be through a local(ish) library or, if you're a student,
through your university.  I believe that most people in the UK can get
it from their local public library.  If you live in California, you
can get it through the Los Angeles Public Library, but you have to go
down there once in person to get a card.  Several other US libraries
also give access to their patrons and have different rules about who
gets to be a patron.

Similarly, your library (or one you can get access to) may give you
access to one of the historical newspaper databases, such as that of
the _New York Times_.  (Make sure that it's actually a *historical*
database, going back to the beginning of the run rather than simply
the last few years.)  There are a few freely-available ones on the
web, such as the _Brooklyn Daily Eagle_.

Finally, Google Books is a wonderful tool, if tricky to use.  It has
works going back (at least) to the sixteenth century.  If you include
in your search terms a date range such as "date:1600-1899" (without
quotes) you'll limit what you get to books from the seventeenth
through nineteenth century.  Mostly.  There are a lot of errors in
dating (often due to OCR errors), so I tend to use three rules: First,
I make sure to select "Limited preview and full view", since otherwise
there's no way to do the next two.  Second, look inside the work for
internal evidence of the date.  Typically, this will be on the title
page, copyright page, or (for periodicals) on the headings of nearby
pages.  (Actually, the first part is to use common sense.  If the
result says the book is from 1799 and talks about the University of
Oregon or gives a date as 1945, it's probably not right.)  And third,
make sure that it looks like the page that has your result looks like
it belongs with the page you found the date evidence on.  Sometimes
more than one book gets smooshed together, and the result is from a
later work.