Information Technology Reference
In-Depth Information
understands the area, establish a research history for the paper, and allow the
interested reader to gain a deeper understanding of the research by following it
from its inception. Gratuitous self-reference, however, undermines these purposes;
it is frustrating for readers to discover that references are not relevant.
On rare occasions it is necessary to refer to a result in an inaccessible paper. For
example, suppose that in 1981 Dawson wrote “Kelly (1959) shows that stable graphs
are closed”, but Kelly (1959) is inaccessible and Dawson (1981) does not give the
details. In your write-up, do not refer directly to Kelly—after all, you can't check
the details yourself, and Dawson may have made a mistake.
According to Dawson (1981), stable graphs have been shown to be closed.
According to Kelly (1959; as quoted by Dawson 1981), stable graphs are closed.
The second form tells readers who originated the result without the effort of obtaining
Dawson first. Kelly's entry in the bibliography should clearly show that the reference
is second-hand.
Regardless of whether you have access to original sources, be careful to attribute
work correctly. For example, some authors have referred to “Knuth's Soundex algo-
rithm”, although Knuth is not the author and the algorithm was at least fifty years
old when Knuth discussed it.
Some readers of a paper will not have access to the publications it cites, and
so may rely on the paper's description of them. For this reason alone you should
describe results from other papers fairly and accurately. Any criticisms should be
based on sound argument. That is, it is acceptable to make reasoned criticisms, and a
careful assessment of past work is of great value because ultimately it is how a paper
is regarded that determines its worth. However, only rarely is it acceptable to offer
opinions, and it is never acceptable to use flattery or scorn. Neither belittle papers,
regardless of your personal opinion of their merits, nor overstate their significance;
and beware of statements that might be interpreted as pejorative.
Robinson's theory suggests that a cycle of handshaking can be eliminated, but
he did not perform experiments to confirm his results [22].
Robinson's theory suggests that a cycle of handshaking can be eliminated [22],
but did not report experimental confirmation.
Robinson's theory suggests that a cycle of handshaking can be eliminated [22],
but as yet there is no experimental confirmation.
Careful wording is needed in these circumstances. When referring to the work of
Robinson, you might write that “Robinson thinks that…”, but this implies that you
believe he is wrong, and has a faint odour of insult; you might write that “Robinson
has shown that…”, but this implies that he is incontrovertibly right; or you might
write that “Robinson has argued that…”, but then should make clear whether you
agree.
A simple method of avoiding such pitfalls is to quote from the reference, partic-
ularly if it contains a short, memorable statement—one or two sentences, say—that
