College and Research Libraries


neophytes in a very technical field, is not 
at all bad. 

It was not the intent of the editors or au-
thors to provide a handbook of procedure, 
statistical or otherwise; their job was rather 
to suggest ways of approaching research, pos-
sible refinements, relevant statistical proce-
dures, all of which are carefully documented 
in extensive notes and footnote references 
to which the interested reader may go for 
further, more detailed information. T o 
apply the cliche "mine of information" would 
be misleading; the volume is rather of the 
nature of a detailed report of the activities 
of an assay office. T h e suggestions of how 
and where to dig are there; the digging the 
student will have to do for himself.—LeRoy 
Charles Merritt, University of California. 

F l o w of Scientific 
Information 

The Flow of Information among Scientists: 
Problems, Opportunities, and Research 
Questions. Prepared by Columbia Univer-
sity, Bureau of Applied Social Research, 
May 1958. New York: T h e Bureau, 1958. 
202p. (mimeographed). 

This pilot study, prepared for the Na-
tional Science Foundation, was undertaken 
to explore the possible contributions of re-
search by interview methods to the problems 
of exchange of scientific information. Its 
purpose was to formulate questions and to 
identify heretofore undefined categories of 
phenomena. Special attention was, there-
fore, devoted to the more obscure of the 
services performed by the scientific com-
munication system, and on the unplanned 
and apparently accidental mechanisms for 
performing them. 

Seventy-seven scientists at one university 
were interviewed, including biochemists, 
chemists, and zoologists in substantially 
equal numbers. T h e interview outline was 
revised continuously during the study and 
its final version is included as an appendix. 
Average interview time was just under two 
hours. T h e sample was so limited as to make 
sophisticated statistical analysis ridiculous, 

so the analysis of the data in the report is 
essentially qualitative and discursive. 

T h e scope set for the study was all the 
channels through which scientists exchange 
and gather information, and all functions 
which scientific communication facilities are 
called upon to perform. 

Since so much emphasis has been placed 
upon means for finding answers to specific 
questions, special emphasis is laid, in this 
study, on instances in which scientists se-
cured answers to specific questions in ways 
other than those designed for this purpose. 
Twenty-eight reports were obtained on in-
formation sought outside the "regular chan-
nels of search," primarily by asking other 
people. Of these about two-thirds dealt with 
details of procedure. A few involved per-
formance of experiments or expert judg-
ments but most of the remaining two-thirds 
were materials of the type that should nor-
mally appear in the literature and about 
half actually did involve asking someone 
else to provide the literature citations. T h e 
first chapter suggests as projects for further 
research: (1) to determine how adequately 
information from personal sources is avail-
able; (2) should more varieties of informa-
tion be securable in print, or should in-
formal channels be made more widely us-
able? (3) how can informal (i.e. personal) 
channels be made more widely usable? (4) 
should more be made available through 
print (a) by having more printed or (b) by 
making what is printed easier to find? (5) 
what makes published information hard to 
locate? (6) why is information of certain 
types seldom published? 

Chapter II, dealing with the problem of 
keeping scientists abreast of current develop-
ments in their specialties, reports only read-
ing and personal contacts, with reading of 
journals in the specialty as the primary tool 
of two-thirds of those reporting. T h e ques-
tions proposed for future research are: (1) 
Does any significant amount of current in-
formation fail to appear in the literature? 
(2) Why are published items missed? (3) In 

what fields are published items most likely 
to be missed? (4) What are the forms of 
personal communication that work? (5) How 
much access do scientists in varying positions 
have to personal communications? (6) What 
clues to pertinence of articles are lacking? 

MARCH 1959 163 



(7) At what level of efficiency is scanning of 
journals done?, etc. 

T h e other functions, covered in Chapter 
I I I , are review of the state of the art, estab-
lishing reliability of a source, redirecting 
attention to different approaches or aspects, 
eliciting feedback from other scientists, help-
ing to assess the position of a topic within 
the current research market. T h e topics for 
research suggested in this chapter are: (1) 
What is incidence of communications serv-
ing the function? (2) How many of these 
involve forms of communication that are 
problematical? (3) How important to sci-
ence were these experiences? (4) How did 
they come about? and (5) What are the 
corresponding experiences of those who 
lacked easy access to these sources? 

Chapter IV, which provides supplemen-
tary comments on the main channels of 
communication, points out that the aver-
age number of journals read is thirteen, 
that biochemical and chemical literature is 
dispersed less than zoological literature. 
Similar miscellaneous notes are included on 
reviews, abstract journals, conferences, and 
personal contacts. 

This study is interesting in that it rep-
resents an attempt by trained social scien-
tists to develop a clearer statement of the 
nature of the problems of scientific com-
munication and to isolate those that might 
profitably be studied. Unfortunately, despite 
the stress laid on isolating non-conventional 
forms of communication, there is little ex-
posed here that is not common knowledge 
among the practitioners in the field of spe-
cial library work or documentation, and the 
program of investigation might have prof-
ited from using the current management 
engineering technique of including in the 
survey team at least one person who is com-
petent in the discipline being investigated. 
T h e fundamental problem in this type of 
approach to the problem of scientific com-
munication is that it attempts to derive re-
liable data from the consensus of a group 
that does not include specialists in the field 
in which they are being queried. This, like 
asking visiting bankers what they think of 
the cooking on railway dining cars, may 
elicit consensus but may not necessarily 
point to valid action. This is best exempli-
fied in the discussion of review journals 

(p. 140-145) in which of fifty-three men 
asked to list the distinguishing features of 
good and bad reviews, twenty-nine failed 
to give any and those that did give features 
for identifying bad reviews couched them 
in such general language as to be meaning-
less in operational terms. 

Probably one of the most important fea-
tures of this report is its recognition of the 
limitations of the method, or perhaps even 
the questions, in view of "the nature of 
specialization among the basic research-
ers. . . ."—Ralph R. Shaw, Graduate School 
of Library Service, Rutgers University. 

Dorking Conference 
Proceedings of the International Study Con-

ference on Classification for Information 
Retrieval. Held at Beatrice Webb House, 
Dorking, England, 13th-17th May, 1957. 
London: Aslib; New York: Pergamon 
Press, 1957. 147p. 

T h e Proceedings of the International Study 
Conference on Classification for Informa-
tion Retrieval, held at Beatrice Webb House, 
Dorking, England, May 13-17, 1957, makes 
widely available the principal addresses, 
discussions, conclusions, and recommenda-
tions of the Conference. T h e forty invited 
participants represented a broad national 
spectrum including France, Germany, India, 
Italy, the Netherlands, Unesco, the Unit-
ed Kingdom, and the United States. Invited 
representatives from the Soviet sphere did 
not attend. 

It seems possible that the long-range sig-
nificance of the Dorking Conference will 
not emanate so much from its own sub-
stantive achievements. T h e limitations of 
conferences of this kind in terms of valid 
research achievement are obvious. But as 
pragmatic devices to create a more conducive 
mental environment for cooperative research 
in needed areas, conferences of this kind 
find a level of reality and usefulness. In this 
sense, the addresses themselves at Dorking 
may be regarded as a kind of "busy" em-
broidery-work around this deeper, pragmatic 
function. They cover a wide range from 
pious generalities, to "chauvinism" concern-
ing a particular system, to highly specialized 

164 COLLEGE AND RESEARCH L I B R A R I E S -