40TH ANNIVERSARY

J M A T E R S C I 4 1 (2 0 0 6 ) 5 9 3 –5 9 6

A review of reviews
R O B E RT W. C A H N
Department of Materials Science and Metallurgy, University of Cambridge,
Pembroke Street, Cambridge CB2 3QZ, UK

The different senses of ‘review’ in the scientific literature are distinguished and their roles are
sketched. Special attention is paid to critical reviews of scientific books and to the role of
referees/reviewers of papers submitted to scientific journals. Some gaps in provision are
pointed out.
C 2006 Springer Science + Business Media, Inc.

1. Introduction (b) A researcher who needs to use an experimental tech-

From its very first issue, just 40 years ago, the Journal nique new to him, or an unfamiliar theoretical approach,
of Materials Science (JMS) has carried review papers, requires a relatively elementary explanation to get started.
covering a huge range of themes. As the first chairman of (c) An inquisitive scientist (probably young) browses,
editors I judged this to be a crucial function of the new hoping to be mentally stimulated by learning about some
journal; in fact, the importance of this function can be recent development, presented very concisely.
judged from the fact that at an early stage review papers (d) A researcher who feels the need for a more substan-
became the responsibility of a dedicated editor. tial treatment of a complex subject than he can get from a
In this short essay, I aim to anatomise the wide variety single paper, proposes to borrow or buy a book (a ‘mono-
of reviews in science in general, and in materials science graph’) and needs a critical evaluation of those recently
and engineering more specifically. published in the field concerned.
To get an idea of just why reviews are so important, (e) A scientist wishes to get a feel for a branch of science
consider the following statistic. In a book [1] published remote from his professional activities, and searches out
in 1999, we are told that the longest index ever printed a popular or semi-popular treatment.
up to that time was the Chemical Abstracts Twelfth Col-
lective Index, 1987–1991, which consists of 115 bound To satisfy these very distinct requirements, the prospec-
volumes with more than 200,000 pages containing over tive reader has a choice between a number of categories
35 million entries. Each of those entries refers to a pub- of review. To be specific:
lished research paper. A conscientiously written chemical
(1) A highly specialised book, or monograph.
review paper can subsume, say, 200 of that 35 million in
(2) A textbook (devoted to a broad subject) which may
such a way that for most scientific purposes, the review
be introductory or advanced. If it is a long multiauthor
can substitute for a trawl through those hundreds. . . or at
text, each chapter may be advanced enough to verge on
the least, the reader of the review will know which few
the character of a monograph.
of those 200 publications he needs to study closely. This
(3) A very thorough and detailed review paper, pub-
is obviously helpful, bearing in mind that most scientific
lished in a journal exclusively dedicated to such reviews,
papers are fated to vanish without trace and never to be
which can include ‘progress’ serials or annual review se-
cited.
rials (such as Progress in Materials Science or Annual
Reviews save us from drowning in information.
Reviews in Materials Science).
(4) A somewhat less detailed review paper, published in
a general archival journal (like JMS, or an ‘overview’ in
2. Motives for consulting reviews; categories of Acta Materialia).
reviews (5) An article in an encyclopedia, general or technical
A research scientist can have a number of quite distinct (like the recent Encyclopedia of Materials: Science and
objectives in scanning or reading reviews: Technology [2]).
(6) An essay, typically around 1000 words in length,
(a) A specialist wishes to find out what is new in his which concisely summarises a recent advance. This essay
specialism. can appear in a newspaper (typically written by its science

0022-2461
C 2006 Springer Science + Business Media, Inc.

DOI: 10.1007/s10853-006-6473-3 593
40TH ANNIVERSARY

editor), a popular science journal, or in a non-archival can be submitted without prior invitation. Provided the ed-
section of a broad-spectrum archival journal like Nature, itors are paragons of wide knowledge, there is something
Science or Materials Today. to be said for the invitation principle. That way, a good
(7) A critical review of a recently published textbook or balance between topics is most readily ensured. However,
monograph, normally found in a scientific journal. There no editor, however learned, can know about all topics
are no critical reviews to be found of categories 3–6. worthy of being reviewed, or about all the new, young
It will be clear that the word ‘review’ has several distinct experts awaiting their opportunities. A mix of invited and
meanings; in particular, a critical review of a book is unsolicited articles is probably best.
something fundamentally different from a review of a
scientific speciality. On top of that. . .
(8) an expert who is invited to adjudicate a paper submit- 4. The essay
ted to a scientific journal is called a reviewer or a referee, The essay that opens the browsing reader’s eye to a recent
as is an expert who is asked to provide his opinion of an scientific development is in a time-honoured tradition,
application for research funds. certainly going back to Victorian times. Such an essay
can be the printed version of what was originally a pub-
Like all the most important words, ‘review’ has many lic lecture, for instance a Friday-evening discourse at the
nuances. Royal Institution in London, or it can be part of the broad
cultural newspaper section which the press in mainland
3. The specialised review Europe is apt to call the ‘feuilleton’. All these are aimed at
As already indicated, these can appear in a variety of loca- the general listener or reader, as distinct from the profes-
tions, and correspondingly the level and completeness of sional scientists. The essays that form part of the ‘News
treatment varies extensively. Normally, the least compre- and Views’ (non-archival) section of Nature (a weekly), or
hensive specialised overview will be found in an encyclo- else feature as ‘columns’ in a more sharply focused pub-
pedia article. I will take just one topic, picked at random, lication such a Materials Today (a monthly), are aimed at
to exemplify this. If a reader wants to learn about crystallo- fellow-scientists.
graphic texture, alternatively called preferred orientation, I happen to have long experience of writing such es-
and its consequences, he can begin with an encyclopedia says, going back to an invitation in 1967 from the editor
(e.g., [2]), and read concise treatments under the titles of Nature to become materials science correspondent to
‘Textures’ or ‘Microtextural Analysis’ (10 pages or less). that journal. (No doubt my editorial involvement with
If that is insufficient for his purpose, he can move on to a JMS, then quite new, and dedicated to what was then a
book chapter, such as my own 50-page chapter, ‘Measure- new field, led to this invitation). In the 38 years since
ment and Control of Textures’ in a multiauthor textbook then, I have written about 100 essays for Nature on a
[3]. If that in turn is not deep enough, the reader can finally great variety of topics loosely included in materials sci-
turn to highly specialised review papers by someone like ence, pitched at a level to appeal to other scientists active
H.-J. Bunge or to an entire book (which in this instance in research. (I was specifically not writing for the gen-
will be over 40 years old and written in German). In each eral non-scientific public. . . their requirements are met in
of these reviews, he will find plentiful references to the other publications). Just recently, I had the opportunity to
next, more advanced level of reading-matter. The starting- do the same, in the form of 24 ‘Cahn’s Columns’, for Ma-
point, in an encyclopedia, can usually be quickly found terials Today. The topics to be covered always came to my
by using the elaborate searching lists provided in such a attention through a settled habit of browsing in a variety
book. Those who feel at ease with them use collections of of sources and picking out topics that seemed to me gen-
abstracts to search for sources; nowadays, search-engines uinely important, or intriguing, or both. Occasionally, an
on the internet are undoubtedly more widely used. unsolicited reprint received through the mail stimulated
Turning from the user to the provider, the topic of a my interest sufficiently to generate an essay.
specialised review varies a great deal. Typically, the title A collection of 100 such columns, not all from Na-
may denote a specific material, a family of materials, a ture, appeared in the form of a book [4]. In the preface,
physical, chemical or mechanical kind of property, an ex- I remarked that a great variety of scientists – geologists,
perimental technique, a processing strategy, a theoretical geophysicists, psychologists (as well as ‘straight’ physi-
approach. My impression is that the topics reviewed in cists and chemists, not to mention materials scientists) had
JMS are quite often unconventional, which enhances their written to me to indicate what they had found of interest
overall level of usefulness. Listing all the review articles in some of these essays, and it was often some incidental
published in recent volumes, as done in JMS, enhances aspect, the broad interest of which had never occurred to
their usefulness. Reviews in category 3, above, published me. I also remarked that “one test of efficacy in getting an
in journals that are exclusively dedicated to reviews, in involved ‘story’ across to readers is whether some of them
some instances have to be commissioned by the editors are deceived into believing that the writer is a specialist,
(preferably, there are several of these), in other instances one who is professionally centred on that week’s topic”.
594
 40TH ANNIVERSARY

One thing is certain: the habit of writing essays of this However, non-scientific periodicals (journals of opinion)
type fertilises the mind of the writer. seem to have no hesitation about paying their reviewers. I
have noticed that people learned in the humanities are apt
to list the book reviews they have written as part of their
5. Book reviews curriculum vitae, whereas very few scientists do this.
This is probably, today, the most neglected form of Some years ago, a publisher with whom I was involved
review. Books of broad popular appeal, accessible to discussed the tentative idea of starting a journal dedicated
non-scientists, do get reviewers’ attention in journals exclusively to critical reviews of scientific books (not, I
of broad coverage, such as Nature or Science, or in- hasten to add, restricted to books published by that firm),
deed in newspapers or magazines. The problem lies with with elaborate provisions to keep possible corruption at
specialised monographs: these are invariably expensive, bay, but in the end they decided that in commercial terms
often very much so, and it is crucial both that con- the project was too problematic. Perhaps it is time to think
stricted library budgets should not be wasted on infe- about this once more.
rior books and that outstanding treatments are widely
recognised for their merits and receive good sales. . .
failing this, publishers are forced on to a trajectory
of ever more outrageous pricing and ever more con- 6. Referees/reviewers
tracting sales, and eventually to the abandonment of I am referring here to the experts who are invited to report
monograph publishing altogether. Careful, expert and to a journal or book editor about the quality of a paper
honest book reviews are essential to the health of sci- or chapter submitted for possible publication. In Britain,
entific publishing. they are known as referees, in America they are more
A problem for journals such as JMS is that organising commonly (but by no means always) called reviewers.
the commissioning and publication of book reviews is a The American term may lead to confusion with other
demanding process and overburdened editors, especially functions, but it does properly denote what an editor’s
the majority of editors who have other professional du- adviser is supposed to do. . .. namely, advise the editor,
ties besides editing, simply keep clear. (When JMS began who decides. ‘Referee’ as a term is reminiscent of the
publication, a few book reviews were included at my urg- man dressed in black who scurries around a football field
ing, but my editorial colleagues soon put their foot down and dominates it. A football referee decides. . . he does
and the practice was discontinued). A few journals have not advise.
recognised the importance for their readership of publish- It is generally accepted that the maintenance of qual-
ing reviews of books at different levels: one such journal ity in a journal, especially one that publishes the results
with which I have been actively involved, Contemporary of original research, is dependent on the input of ref-
Physics, devoted primarily to specialised review articles erees/reviewers. This has gone so far that some schol-
in sense (3), above, also has a dedicated book reviews ars, in India particularly, divide their list of publications
editor and about half of each issue is devoted to a very into those that were published in refereed journals and
wide range of critical reviews of books about all aspects those published elsewhere. Some years ago, John Ziman
of physics (including physics of materials). This takes a (a physicist turned into a sociologist of science) remarked
lot of editorial time and care but the result is eminently “An article in a reputable journal does not merely repre-
worthwhile. sent the opinion of its author; it bears the imprimatur of
The house journals of many professional societies do scientific authenticity, as given to it by the editor and the
recognise the importance of publishing book reviews but referees he may have consulted. The referee is the lynch-
the number of pages devoted to this is usually very modest. pin about which the whole business of Science is pivoted”
Some journals at the borders of science and the humani- [5].
ties, such as Isis, have extensive book review sections, like The end-result of the combined critique applied by the
Contemporary Physics, but usually few books of ‘hard’ referee and the editor appears to be very dependent on the
science are covered here. I believe that there is no alterna- discipline. A study in 1971 by the renowned sociologist
tive to the inclusion of book reviews in the major archival of science, Robert Merton, and his collaborator Harriet
journals, such as JMS. The publishers of such journals Zuckerman [6] revealed that whereas in a selection of
should subsidise the extra costs of undertaking this func- physics and chemistry journals the average rejection rate
tion, because after all in the long run this must necessarily of papers ran at 24 and 31% respectively, in sociology,
benefit their book-publishing branches. language/literature and history it ran at 78, 86 and 90%
Most book reviewers are rewarded by being allowed to respectively! Nobody has made such a study for materi-
keep the book (though I have encountered a professional als science journals. In the same survey, we learn that in
society which asked for the book back for incorporation the Physical Review, 72% of references are to papers pub-
in its library!). Payment is rare: only a journal with a lished within the preceding five years, whereas in journals
very large circulation apparently can afford such a policy. devoted to history and art criticism, only 10–20% of ref-
595
40TH ANNIVERSARY

erences are to such recent sources. Being a referee for areas of scientific publishing which are untouched by the
a historical journal must be quite demanding: after all, exercise of judgment. In particular, this little paper apart,
one of a referee’s explicit functions is to check whether there are no such things as reviews of reviews, and more
appropriate reference has been made to earlier work. particularly, there are no comparative assessments of the
The specific role of the referee/reviewer for a scientific merits of different surveys of (more or less) the same
journal has not received very extensive analysis, though topic, taking monographs, book chapters, review papers
from time to time an investigator will submit the same and encyclopedia articles together. This lack requires
manuscript to several referees to find out what measure of attention from the community of editors and publishers:
agreement results. The result can be disconcerting. The materials scientists may as well exercise leadership here.
most detailed general analysis that I know of is in the pro-
ceedings [7] (including very extensive discussion) of a
symposium held at Vanderbilt University in Tennessee in References
1992, under the title Editing the Refereed Scientific Jour- 1. H .- J . Q UA D B E C K – S E E G E R , “World Records in Chemistry”
nal. In that book, I have a personal paper about the proper (Wiley-VCH, Weinheim, 1999) p. 138.
2. K . H . J . B U S C H OW, R . W. C A H N , M . C . F L E M I N G S ,
role of referees, warning of the dangers of the notion that B . I L S C H N E R , E . J . K R A M E R and S . M A H A JA N (editors-in-
a referee should essentially have the last word in deciding chief), “Encyclopedia of Materials: Science and Technology” (Elsevier,
whether or not a paper is acceptable. The usefulness of Oxford, 2001).
that symposium to a reader who is a materials scientist is 3. R . W. C A H N , ‘Measurement and Control of Texture’, in “Processing
however somewhat limited by the strong emphasis on the of Metals and Alloys” edited by R. W. Cahn (VCH, Weinheim, 1991)
p. 429.
biomedical literature among the participants. That cluster 4. Idem., “Artifice and Artefacts: 100 Essays in Materials Sci-
of disciplines has its own very specific problems. ence” (Institute of Physics Publishing, Bristol and Phildelphia,
1992).
5. J . M . Z I M A N , “Public Knowledge : The Social Dimension
7. What next? of Science” (Cambridge University Press, Cambridge, 1966) p.
The reader will have noted by now that I support the 148.
exercise of expert critical judgment in the scientific 6. H . Z U C K E R M A N and R . K . M E RT O N , Minerva 9/1 (1971)
66.
literature, even though there are sociologists in plenty
7. R . A . W E E K E S and D . I . K I N S E R (editors), “Editing the Refereed
who have no use for judgment: ‘opinion’, in their view, is Scientific Journal: Practical, Political and Ethical Issues” (IEEE Press,
deeply suspect as a basis for decisions. There are indeed New York, 1994).

596