Andreas van Cranenburgh
	November 10, 2010
	Mechanisms of Meaning, exercise 3

1.1 For a given pair of words, a range of words is returned ordered by
the strength of their association with the two given words. So the
beginning of the range is most strongly associated with the first word,
and the end most strongly with the second word.

1.2 It is very clear that utterly has a pejorative, negative connotation,
whereas absolutely has a positive connotation. This connotation must
affect their meaning and guide their usage.

1.3 The pair war - conflict. `War' is a standard term, `conflict' is a 
diplomatic euphemism. This is reflected in their collocates. For example `war'
has collocates such as prisoners, fought, killed. But `conflict' has such terms
as situations, potential, resolution. It also seems that the word war is used
for historical situations (falklands, vietnam, korean are strong collocates),
whereas conflict is used in more diplomatic situations when a possible future
war is being discussed.

2.1 Clearly in fiction the occurrences of chair are about furniture, because
they are referred to as physical objects (nearby chair, straight chair, heavy
chair, etc.). In academic literature chair refers to a position or an office,
e.g., vice chair, chair of the national [...], past chair). The conclusion is
that chair is ambiguous between the physical furniture meaning, and the
institutional title meaning.

2.2
Comparing collocates for `wine' in spoken and fiction sections. The first
striking feature is that in the spoken section, there are only 11 collocates,
whereas in the fiction section there are as much as 54.
Furthermore it is striking that the six most strongly associated collocates
for the spoken section, namely `sparkling,' `french,' `great,' `other,' `nice'
and `little,' are the lowest ranked for the fiction section -- perhaps in
fiction the writers try to avoid cliches from spoken language.

Comparing collocates for `evidence' in academic and fiction sections.
In the academic sections there are lots of technical collocates, such as
`significant,' `experimental' and `clinical.' In the fiction section the
collocates include collocates known from court rooms such as `incriminating,'
`forensic' and `circumstancial,' as well as general collocates such as `hard,'
`physical' and `incontrovertible,' but not technical words as in the academic
section.  It can be concluded that the use of the word evidence is more
specialised in the academic section.

3.1
The adjectival collocates of the word `quite' compared in 1870 and 2000.
I had previously read that the meaning of `quite' has changed, so I decided
to investigate. It seems that in 1870, quite was used with the sense of
`completely', whereas in 2000 it also has the sense `to a large extent', in
other words as a degree modifier. Compare a sample from strong collocates: 

1870: quite unconscious, unknown, indifferent 
2000: quite funny, normal, amazing.

It also seems that the collocates from 1870, at least the top 20, are negative
or neutral, whereas in 2000 there are collocates with very positive meanings as
well, e.g., beautiful, extraordinary, powerful. Either way, it is striking that
at first glance there seems to be no overlap at all in the two lists of
collocates.

3.2
a) the word `database.' With all the exponential developments in technology it
is perhaps too easy to find a neologism that has increased until the present
time.

Figures in words per million, 
1800-1970: 0
1980: 0.83 
1990: 6.87
2000: 8.66

Consequently, and not surprisingly, the frequency is monotonically increasing.

b) the word `celerity'. I tried to think of an old-fashioned word and
somehow this word popped up in my head.

Some sample data points of words per million,
1820: 5.20
1870: 2.52
1930: 0.98
2000: 0.03

This word, meaning quickness or rapidity has become very rare nowadays,
whereas it used to be quite common. The frequency of this word is clearly
decreasing monotonically.