The paper investigates the problems of using commercial search engines for
web-link research and attempts to clarify the nature of the problems involv
ed in the use of these engines. The research finds that search engines are
highly variable in the results they produce, are limited in the search func
tions they offer, have poorly and/or incorrectly documented functions, use
search logics that are opaque, and change the search functions they offer o
ver time. The limitations which are inherent in commercial search engines s
hould cause researchers to have reservations about any conclusions that rel
y on these tools as primary data-gathering instruments. The shortcomings ar
e market-driven rather than inherent properties of the web or of web-search
ing technologies. Improved functionalities are within the technical capabil
ities of search engine programmers and could be made available to the resea
rch community. The findings also offer mild support for the validity of the
connection between web links and citations as analogues of intellectual li
nkage.