ChapterPDF Available

Information’s Magic Numbers: The Numerology of Information Science

David Bawden and Lyn Robinson
Information’s Magic Numbers:
The Numerology of Information Science
 Introduction
Two themes were presented at the Libraries in the Digital Age (LIDA) conference,
held in Zadar, Croatia, in June 2014. The rst, chaired by David Bawden, focused
on “qualitative assessment”. Blaise Cronin chaired the second theme, focused on
“altmetrics”. They, as well as several other conference speakers, emphasized the
complementary nature of qualitative and quantitative methods; while quantita-
tive data is of unarguable importance, it must be interpreted insightfully and used
with care.¹ Applying numbers sensibly has always been a major concern for Blaise
Cronin, as is attested by his publication list. His academic webpage² notes that
“much of his research focuses on collaboration in science, scholarly communi-
cation, citation analysis, the academic reward system and cybermetrics—the
intersection of information science and social studies of science”, while his
Wikipedia entry³ describes him as being jointly an “information scientist and
bibliometrician”. Despite this strong informetric focus, Cronin has had a long-
standing concern about the potential descent of this aspect of the information
science discipline into a “new age of numerology,” due to over-use and misuse
and of bibliometrics and altmetrics; see, for example, Cronin (1998; 2000), Cronin
and Sugimoto (2015), and Priego (2012). It is therefore appropriate to include in
this volume a chapter on the numerology of information science; to ask to what
extent we are able to identify a few numbers which may helpfully encapsulate
important aspects of the subject.
Numerology, roughly the belief that numbers in general, and integers in par-
ticular, have their own nature and properties, and can of themselves inuence
events, is rather out of favor nowadays, being regarded as a pseudoscience. The
impeccable scientic belief that the regularities of nature can be captured by sim-
ple mathematical relationships is a long way from Blair’s (1976, p. 81) notion that
“numbers, quite distinct from their empirical use, become a language, as full of
1The presentations may be found on the conferencewebsite at
DOI 10.1515/9783110308464-002, © 2020 Cassidy R>. Sugimoto, published by De Gruyter.
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 License.
DOI 10.1515/9783110308464-012, © 2020 David Bawden, Lyn Robinson, published by De Gruyter.
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 License.
Information’s Magic Numbers: The Numerology of Information Science | 181
metaphor and dimension as poetry”. However, before sneering at the idea that
numbers in themselves can have a signicance, we should remember that the
long-standing, and still inuential, Platonic tradition within science views num-
bers as having their own objective existence, and indeed that the physical uni-
verse and everything in it is, at root, a mathematical structure made of numbers
(see Tegmark [2014] for a recent and accessible account of this position).
As well as numbers per se, numerology is often taken, usually critically, to
mean an enthusiasm for simple numerical formulae, usually involving integers,
capturing some signicant aspect of reality. These have been seen in both the sci-
ences and the social sciences: notoriously, the British physicist and astronomer
Sir Arthur Eddington spent many years seeking simple integer relationships as the
clue to the universe (Kilmister, 2005). It is clear that there are strong relations be-
tween numbers, the physical world and cultural issues, as is clearly shown by the
sequence of “kissing numbers”, the number of spheres which in any space exactly
bound a further identical sphere (Weisstein, n.d.); two points on a one dimen-
sional line bound a third point, six circles circumscribe a seventh, and twelve balls
circumscribe a thirteenth. The resultant sequence of “kissing number plus one”—
three, seven, thirteen—captures the principal signicant/lucky/unlucky numbers
in numerous cultures, and is numerologically present in the ‘leader with twelve
followers’ meme of Christ, Osiris, King Arthur, and others (Blair, 1976).⁴ Therefore,
despite the dangers of slipping into a facile numerology, simple numbers and in-
teger relations may still be worth investigating.
There are, in fact, relatively few such simple numbers and number relations
in information science, and what exists was imported from adjacent disciplines.
In truth, they are not all very simple: one is very large, some have alternatives, one
is a sequence, and one is innite. These numbers encapsulate a variety of issues:
how much information there is, or could be; the optimal size of communicating
groups; the structure of information networks; the distribution of information ac-
tivities; and the limits to the growth of knowledge. We nd that sometimes, but
not always, the actual number is less important than the theoretical perspective
to which it points. We begin by considering the big picture: how much information
there is, or could be, in the human context and in the universe. Then we move to
the smallest scale, the information associated with the conscious attention of a
single person. From there, we move up the scale, to information associated with
4It would have been nice if the four-dimensional kissing number, which was not known until
2003 (Pfender & Ziegler, 2004) and cannot be intuitively grasped like the small dimension equiv-
alents, had also related to some culturally signicant number. Disappointingly, it was shown to
be 24, and 25 does not appear to have signicance in any culture.
182 | David Bawden and Lyn Robinson
groups, with networks, and with disciplines; and nally up to the largest scale, to
the innity of possible recorded information.
The Universal Number: 8 × 1021
The most fundamental number-related question we can ask about information
is simply: How much information is there? This leads to a spin-o question: How
much information could there be? Both, perhaps not surprisingly, are dicult to
answer accurately. Attempts to answer such questions have been reviewed by
Bawden and Robinson (2012a), Gleick (2011), Davis and Shaw (2011), and Floridi
Before the advent of widespread digital information, the “How much informa-
tion is there?” question was generally answered in terms of counts of documents:
how many books, articles, reports, etc. had been published. For example, Jinha
(2010) suggested that the total number of scholarly articles had reached fty mil-
lion. More recent attempts have had to include the much larger amount of born-
digital information—an intrinsically more dicult process—with results that can
only be approximate.
The rst attempt in the digital era to address this question in a rigorous way
was the “How much information?” study from the School of Information Man-
agement and Systems at University of California Berkeley, United States (U.S.),
rst carried out in 2000, and repeated in 2003 (SIMS, 2003). The study estimated
that approximately 12 exabytes of information had been recorded by humanity
before the general use of computers, but was being dwarfed by the amounts now
being generated and stored. About 5 exabytes of new information was stored dur-
ing 2002; equal to 37,000 times the information in the Library of Congress, or
800 megabytes/30 feet (10 metres) of bookshelf content per person on the planet.
However, more than three times this amount of information was communicated
through electronic channels but never stored.
Later studies (Hilbert & Lopez 2011; Ganz & Reinsel 2011; Hilbert, 2014) have
suggested that the amount of information broadcast each year, in increasingly
varied formats, was approaching 2 zetabytes by 2007, that current capacity of all
information storage devices approaches 300 exabytes, and that the total grew to
over 1600 exabytes between 2007 and 2011. Floridi (2014, p. 13) points out that this
amounts to enough information being generated each day to ll all the libraries
in the U.S. eight times over, and that the gure is likely to grow threefold every
four years, so that there may be expected to be 8 zetabytes (8 times 1021 bytes) of
information by 2015; this gure is taken as ‘the number’ for this section.
Information’s Magic Numbers: The Numerology of Information Science | 183
One consequence of this, as Floridi (2014) points out, is that since 2007 in-
formation has been produced at a faster rate than have storage devices to handle
it; this despite Kryder’s Law, which shows that the capacity of storage devices is
increasing at an even faster rate than is processing capability, the latter obeying
Moore’s Law. That we cannot therefore store all our information arguably does
not, in fact, matter; the great proportion now is data generated by machines and
used by machines, without any need for longer-term storage for human interven-
tion or reection.
The actual value of these numbers is immaterial—what is of importance is
their scale and order of magnitude, and the ways in which they are changing, to
create what Floridi (2014) terms the “infosphere,” an entirely new form of infor-
mation environment.
The second question, how much information could there be, is answered by
considering the capacity of the physical universe to hold bits of information; a
rough estimate, subject to many approximations, is reported by Gleick (2011) to
be about 1090 bits. This gure, though of no practical signicance for information
science, reminds us that information is always physically instantiated, and its pro-
cessing is limited by the constraints of the physical universe. Of course, there are
those who would go further, and say that information is physical per se, but that is
a topic for a dierent discussion; see, for instance, Bawden and Robinson (2013),
the contributors to Davies and Gregersen (2010), and Hjørland (2007).
The Personal Number: 7 (or 4)
One of the most cited papers in the human sciences is American psychologist
George Miller’s “Magical number 7, plus or minus 2” (Miller, 1956).This paper drew
attention to the signicance of the number in human information processing. The
main nding was that the number of concepts of items which an adult can hold in
conscious attention, or short-term memory, at one time is about seven. Miller, an
early enthusiast for the application of information theory in psychology, used in-
formation theory concepts to speculate about what this meant for the mechanism
of memory. This limit has been widely tested, and was generally regarded as cor-
rect. However, more recent studies, summarized by Cowan (2001), have suggested
that the limit may be lower, between three and ve rather than between ve and
nine. On the basis of these ndings, Cowan recommends a “magical number 4”.
Whatever the exact number may be, it is clearly small, and has implications
for the way in which information is handled and should be presented. However,
there seems to have been relatively little explicit recognition of this in information
184 | David Bawden and Lyn Robinson
science. It is tempting to ascribe to this factor the well-known tendency for users of
search engines to attend only to the items presented on the rst page. This seems
likely to be more a matter of disinclination to spend the time necessary to consider
more items, rather than an inability to hold them in conscious attention at once;
but it may be that some underlying mechanism, associatedwith the number limit,
accounts for both factors.
Knowledge organization systems appear to respect this feature. Decimal clas-
sications may be guided to their ten main sections by a desire for a pleasing
notation, and others, most notably the Library of Congress Classication, follow
the twenty six letters of the Roman alphabet. But the general tendency, follow-
ing Ranganathan’s ve fundamental facets (Hedden, 2010; Broughton, 2006), to
have between four and ten main sections or facets in the great majority of tax-
onomies and thesauri may be seen as an unconscious recognition that this is a
number which enables the user, or at least the compiler, to hold the whole struc-
ture in mind. Miller’s number holds up at more detailed levels of taxonomydesign:
A popular rule of thumb is to go only three levels deep and have only six to eight
concepts per level. These numbers are based on user experience tests, which have
shown that users have the patience to click down only to a third level and can
scan only six to eight term entries at once” (Hedden, 2010, p. 236). This, of course,
reects the similar experience with search engines noted above.
The Group Number: 150
The group number stems from the work of the British evolutionary psychologist
Robin Dunbar, initially inspired by the study of the correlation between brain size
and the size of social groups in primates. This led to the idea that there is a natu-
ral group size for humans; stable communicative relationships can be maintained
with about 150 people (Dunbar, 1993; Dunbar, 2008; Dunbar, 2012). The number
derived from the correlation was actually 148, but it has generally, and sensibly,
been rounded to 150. This idea is of evident importance for information science,
since it is well-known that close acquaintances are a major source of information
in most, if not all, contexts (Case, 2012). Further, shared knowledge is a major fac-
tor in the maintenance of social groups (Dunbar, 2012; McPherson, Smith-Lovin, &
Cook, 2001).
Dunbar argues that the size of the group of communicative relationships
which can be maintained at any one time is constrained in part by cognitive
factors, and hence ultimately by some aspect of brain size and structure, and
partly by available time. Direct evidence in humans is provided by correlations
Information’s Magic Numbers: The Numerology of Information Science | 185
of individual dierences in social network size and volumes of social cognition
areas in the cortex and amygdala brain structures (Kanai, Bahrami, Roylance, &
Rees, 2012; Bickart, Wright, Dauto, Dickerson, & Feldman Barrett, 2011; Pow-
ell, Lewis, Roberts, García-Fiñana, & Dunbar, 2012). Empirically, this has been
tested by the observations of groupings in a wide variety of contexts, including
hunter-gatherers, farming communities, military formations, industrial and com-
mercial workforces, Christmas card lists, online social networks, and academic
disciplines (Dunbar, 1992; Dunbar, 2008; Hill & Dunbar, 2003; Roberts, Dunbar,
Pollet, & Ruppens, 2009).
Objections to the 150 value have been raised by those who argue that group-
ings of around 30–50 people are commonly found in hunter-gatherer populations,
arguably the most “natural” form of human grouping (de Ruiter, Weston, & Lyon,
2011). Others have suggested that the number must be much larger because of the
evidence that many people have several hundred contacts on social media (Well-
man, 2012). However, as Dunbar (2012, p. 2195) asserts, “there is now considerable
evidence that groupings of this size [around 150 individuals] occur frequently in
human social organization, and that this is the normative limit on the size of per-
sonal social networks among adults.”
A more nuanced viewpoint, rather than seeking to insist on a single number
to encapsulate the complexities of social interaction, is to see a series of num-
bers, reecting dierent strengths of social ties, and of shared knowledge and
perspectives. These groups exhibit the kind of “small world” network structure
and behavior which will discussed later. Typically, these represent groups with
rough sizes 5, 15, 50, 150, 500, 1500, which can be seen as circles, each includ-
ing those inner, with a scaling factor of about three (Dunbar, 2012; Hamilton,
Milne, Walker, Burger, & Brown, 2007; Roberts, Dunbar, Pollet, & Ruppens, 2009;
Zhou, Sornette, Hill, & Dunbar, 2005). Groups of these sizes may be characterized
roughly as follows:
5: a core social group, or “support clique,” to whom an individual would refer
very frequently for support, assistance, information and advice;
15: a “sympathy group,” with whom there are special ties and frequent con-
50: typically a temporary grouping, formed for a particular period or task;
150: the stable inter-communicating group, with regular interaction and
knowledge sharing;
500: the “megaband,” again typically a temporary or pragmatic grouping;
1500: the “tribe”—acquaintances at best, with whom any relationship, or
communication of information, is typically one-way, and there is no little or
no sharing of knowledge.
186 | David Bawden and Lyn Robinson
Of these, the 5 and 150 levels seem particularly signicant: 150 for the reasons
set out by Dunbar, supported by a good deal of evidence, and in keeping with
Shirky’s (2003, n.p.) recommendation for eective online group size (i.e., “larger
than a dozen, smaller than a few hundred”); ve because it appears to be a nat-
ural small group equivalent, related to the idea that spontaneous conversation
and information sharing almost always occurs in groups of not more than four
individuals (Dunbar, Duncan, & Nettle, 1995).
It seems evident that an understanding of this group interaction structure, if
indeed it is valid and omnipresent, is important for several areas within informa-
tion science, perhaps most notably in knowledge management. However, there
seems to have been little examination of the signicance of this group structure
with respect to the communication of information. Studies have established that
the smaller, and more information-intensive and knowledge-sharing, groups re-
quire an investment of time, and ideally substantial face-to-face contact, if their
members are not to slip into the larger, and less eective groupings (Dunbar, 2012;
Roberts, Dunbar, Pollet, & Ruppens, 2009). This seems to be a warning against
reliance on purely digital information sharing, particularly with an assumption
that its scale can be increased by technological means, and typies the value that
such theoretical concepts can bring to information practice.
The Linking Number: 6
The idea that everyone in the world is connected to everyone else by no more
than “six degrees of separation” has become entrenched in popular conscious-
ness through newspaper and magazine articles, plays, TV series, lms, and games
(Six degrees of separation, n.d.). The concept was introduced by the Hungarian
writer Frigyes Karinthy (1929), in his short story Láncszemek (Chains), but became
well-known only with the classic paper of American psychologist Stanley Milgram
(1967). This initiated a research program in what became known as “small world”
phenomena; for a detailed review, from Karinthy onwards, see Schnettler (2009).
In Milgram’s study, randomly chosen participants in the Midwest (U.S.), were
asked to try to send a printed message to a target in New England (U.S.), by send-
ing it to a person with whom they were personally acquainted, asking that it be
forwarded in the same way. Only about 30 % succeeded, and those that did varied
between two and ten intermediaries, with a median of ve. This was the basis for
the idea of “six degrees of separation,” although Milgram did not use this phrase
in his paper. Focusing on the number of nodes, rather than links, in the chain,
he wrote of “ve circles of acquaintances” (Milgram, 1967, p.65). The rather more
Information’s Magic Numbers: The Numerology of Information Science | 187
memorable “six degrees of separation” phrase was introduced two decades later,
in a play with that name (Guare, 1990).
Some limited empirical research in the social sciences investigated this idea
over the next thirty years, until the subject was revitalized by formal mathematical
modeling of network connectivity in all kinds of contexts, not just social (Cal-
darelli & Catanzaro, 2012; Mitchell 2009; Schnettler, 2009). The formal model-
ing results tend to support empirical studies in various contexts, in conrming
commonly occurring short paths through extensive networks, though they do not
support the idea that there is anything special about the number six (or ve); me-
dian chain lengths can vary from three to fteen, according to the nature of the
network. However, a study aiming to replicate Milgram’s work on a much larger
scale using e-mail gave quite similar results, of between ve and seven steps for
the minority of messages which were completed, suggesting that this may be a
natural scale for social information networks (Dodds, Muhamad, & Watts, 2003).
As Stock and Stock (2013, p.384–385) note, the “six degrees of separation”
concept has become synonymous with the idea of “small worlds”. This expresses,
in the social context, the idea that “people are not only linked to their immediate
friends, family, and acquaintances, but they are embedded in a larger structure of
direct and indirect contacts” (Schnettler, 2009, p. 166). More formally, the “small-
world eect” denotes the fact that most nodes in most networks are joined by
relatively small paths; a specic “small-world network” has been identied as one
with a structure intermediate between highly regular and totally random, with
nodes highly clustered, as in regular graphs, and yet with a short path length be-
tween any two nodes, as is typical in random graphs (Schnettler, 2009; Watts &
Strogatz, 1998).
However, despite this theoretical support for short paths, empirical work on
social networks, typically carried out in the sociology domain, have tended to
show that, although extended chains of social contacts were available, they were
used infrequently for nding information (Schnettler, 2009). For example, in a
study of how people found information about job prospects, most used one in-
termediary, or none, and no chain was more than four links (Granovetter, 1995).
The only example of longer chains, with up to nine links and a median of ve,
was found in a study of women in the U.S. seeking a doctor willing to perform an
abortion at a time when legal abortion was severely restricted (Lee, 1969).
Björneborn and Ingwersen (2001, p.74) noted that small world metrics were
potentially relevant to several topics within information science including webo-
metrics, citation analysis, semantic networks, and thesauri, but that there was a
lack of research in these areas. Since then there has been some usage in webo-
metric studies, a typical example being the demonstration that the typical path
link between sites in the United Kingdom (U.K.) academic web network is three
188 | David Bawden and Lyn Robinson
or four (Björneborn, 2006), and in bibliometrics, for example, a study of the co-
occurrence of keywords in databases, where the number reects the distance be-
tween papers measured by the keywords in common (Zhu, Wang, Hassan & Had-
dawy, 2013). The only specic mentions of the “six degrees” idea in the recent
information science literature appears to be James’ (2006) reections on the rel-
evance of the idea to information literacy instruction, and Dennie and Cuccia’s
(2014) application to a chemical literature search assignment.
While considerable research has been carried out within information science
using the “small worlds theory”, this has largely been detailed qualitative studies
of information interactions between groups and networks in limited spaces, phys-
ical or virtual (Savolainen, 2009). Concepts such as ‘density’ from network theory
may be applied (see, for example, Huotari & Chatman, 2001), but generally in
an informal and semi-quantitative way. Even within these caveats, Schultz-Jones
(2009, p.626) found that “library and information service settings [are] a largely
undeveloped context for the application of social network theory and social net-
work analysis.” It may be that there is scope for better integration of qualitative
and quantitative methods, as Schnettler (2009) advocates in general for small
world research, and for a greater focus on contexts closer to our own (disciplinary)
home. The number, whether it be 6 or not, is not, in this case, as important as the
“network thinking” (Mitchell 2009) to which it points.
Finally, we might note that the “six degrees of separation” idea has launched
metrics such as the Bacon number, the closeness of the Hollywood actor Kevin
Bacon to any other actor, based on the actors who have worked with actors who
have worked with Kevin Bacon, and, perhaps more seriously, the Erdös number,
based on how many links of co-authorship link anyone to the Hungarian mathe-
matician Paul Erdös (Grossman, 2014). Perhaps we should establish an analogous
Cronin number: one of us [DB] would be 2, since he has not co-authored with
Cronin, but has co-authored with at least one person who has, LR would have
a Cronin number of 3, on the same basis.
The Network Number: 59
As we have just seen, experiments have shown that messages across small world
networks fail to get through a majority of the time. This may be due to a variety
of context-specic factors, depending on the nature of the network, and the pat-
tern and strength of its connections (Dodds, Muhamad & Watts, 2003; Schnettler,
2009). Milgram (1967) noted a specic, and fairly obvious, point that two groups
within a network may be cut o, if there is no link path joining them, so that there
Information’s Magic Numbers: The Numerology of Information Science | 189
is no possibility of information passing between them. Mathematical analysis of
networks by the American complexity scientist Stuart Kauman has shed an in-
teresting light on network behavior in this respect.
Kauman has shown that, for any network of nodes which are all initially
isolated, adding links randomly between nodes causes a pattern of connections to
build up, steadily and linearly, so that linked groups are created within the overall
network. This may be seen as an instantiation of Ramsey theory, which posits the
unavoidable emergence of regularity in large structures, such as networks. It is
often expressed as the ‘party problem’; how many guests must be invited to a party
(or people invited to link to a social media site) so that a minimum number (the
“Ramsey Number”) will know each other (Gould, nd).
Kauman shows that when 59% of the nodes are linked to at leastone other,
the pattern suddenly and dramatically changes, and the great majority of the
nodes are connected. This is referred to as a network phase transition. An accessi-
ble account of the phenomenon is given by Kauman (1996), and its signicance
is described for computer networks, such as the World Wide Web (Tetlow, 2007)
and for social networks such as the nancial system (Beinhocker, 2006).
The importance of this number for information science is that it should in-
still an awareness that the behavior of information networks of all kinds may
change, suddenly and dramatically, as their interconnectivity increases. It is easy
to assume that overall connectivity within a network, and hence the ability to
pass information between any two of its nodes, will increase in a regular man-
ner, as more individual interconnections are added, depending on the number
of connected nodes. This is the basis for “laws” relating the value of a network,
specically a computer network, to the number of nodes connected. Metcalfe’s
Law, for example, states that the value of a network increases as the square of
the numbers of nodes connected (Floridi, 2014), while a variant due to Briscoe,
Odlyzko, and Tilly (2006) argue for a less rapid growth of n(log n), with n nodes
connected. Kauman’s number shows us that this kind of continuous growth in
network value is only valid up to a point. Beyond this, rather precisely speciable,
point, a qualitative change in the nature, and value, of the network occurs, leading
to an essentially new information environment.
The Distribution Numbers: 90, 9, and 1
The numbers 90-9-1 have been found to represent the distribution of activity
among users of social media sites, including microblogs, such as Twitter, and
wikis, most notably Wikipedia. For every regular contributor, or “superuser”,
190 | David Bawden and Lyn Robinson
there are nine occasional contributors, and ninety “lurkers”, who take informa-
tion but do not contribute with any regularity; as an example, see van Mierlo
(2014). This is an instantiation of a very widespread distribution in information
areas. From our days as information practitioners, we recall it being an article of
faith, stated anecdotally though never written down so far as we know, that in
any complex search for information requiring high recall it was easy to get 90%
of the material, very dicult to get 99% and impossible to get 100%.
These are examples of the ubiquitous power law distributions that govern
the information world, including those of Bradford, Lotka, Pareto, and Zipf (Baw-
den & Robinson 2012a; Egghe, 2005; Rousseau, 2010). As such, they are better
known within the information science community than the other numbers de-
scribed in this chapter, and need less exposition. An appreciation of these laws,
and the numbers which come from them, informs practice in areas such as col-
lection management, information retrieval, institutional bibliometrics, and the
assessment of impact of social media; see, as examples, Corby (2003), Nicolaisen
and Hjørland (2007), Bhavnani and Peck (2010), Åström and Hansson (2012), and
Homan and Doucette (2012). These are thus among the few “magic numbers”
which are used widely and directly in the practice of the information disciplines,
and particularly in scientometrics.
The Knowledge Number: ∞
The knowledge number is generally termed the Champernowne number, after the
British mathematician David Champernowne, who derived and published it while
still an undergraduate student before going on to a career as an economics pro-
fessor (Champernowne, 1933; Pickover, 2012, p. 364–365). While he derived his
number simply as a mathematical curiosity, it has interesting implications for the
information world (von Baeyer, 2003, p. 101102).
We rst choose a base for our number, say binary or decimal. Then we enumer-
ate all the symbols that constitute that number set, then all the pairs, then all the
triplets, and so on, for as long as we wish. In decimal base 10, as Champernowne
originally presented it, we would write 0.12345678910111213141516… or, in the bi-
nary system, we would write 0 1 00 01 10 11 000 001 010 100 … Since we can always
continue adding to this number, it must necessarily be innite in magnitude.
Then we choose a code to convert the number to characters—something like
ASCII or Unicode—and convert our potentially innite number to a potentially
long innite text string. In this innite character string there will be found ev-
erything that has ever been written using the chosen character set, embedded in
Information’s Magic Numbers: The Numerology of Information Science | 191
the (literally) innitely larger set of everything could be written. We will nd the
text of Shakespeare’s Midsummer Night’s Dream, in all its editions, in all possible
languages, and with all possible misprints and errors. We will nd a copy of this
paper, with all these variants, and a copy of all the works which Blaise Cronin has
written, or might have written. This is an instantiation of Borges’ (1998) Library of
[Whose] bookshelves contain all possible combinations of [symbols] – that is, all that is able
to be expressed, in every language. All – the detailed history of the future, the autobiogra-
phies of the archangels, the faithful catalog of the Library, thousands and thousands of false
catalogs, the proof of the falsity of those false catalogs, a proof of the falsity of the true cata-
log, the Gnostic gospel of Basilides, the commentary upon that gospel, the commentary on
the commentary on that gospel, the true story of your death, the translation of every book
into every language, the interpolations of every book into all books, the treatise Bede could
have written (but did not) on the mythology of the Saxon people, the lost books of Tacitus.
(Borges, 1998, p. 115)
And Champernowne gives us this in a number.
The number is of no practical value, but it is a striking formal indication of the
idea that creativity, and growth of knowledge, are unlimited. While our rst num-
ber indicated that the amount of information that can be held within the physical
universe must be nite, creativity is unlimited, and knowledge can grow inde-
nitely (see, for example, Deutsch, 2011; Kauman, 2010).
 Conclusions
It is dicult to state concisely where these numbers t into our understanding of
the information world, and more specically in our understanding of informetrics
and scholarly communication; though it would be dicult to deny their potential
signicance. All these numbers are interesting, and some are of immediate use
for practice; they take us into the areas where the information sciences overlap
with the human sciences, especially psychology, with the physical sciences, and
even with philosophy. It is not evident that there is any metatheory which could
encapsulate them all, and it may be unrealistic to think of anything of the sort.
However, the links between the numbers, for example between Dunbar’s social
groups and Milgram’s small world networks, may serve as a basis for building a
modest theoretical framework.
It is still more unrealistic to seek for a single magic number for information.
Though, if we had to do so, it would probably be 5, since this appears in several
contexts, including cognitive scope, small world links, and optimal group size for
192 | David Bawden and Lyn Robinson
information exchange. The numbers measure attributes of people and groups,
cognition and networks, collections and activities; all three of Popper’s Worlds,
for those who like that ontology as a basic for the subject (Bawden, 2002; Baw-
den & Robinson, 2012B).
The numbers themselves appear rather uid, and usually their exact value
does not matter. It is the general magnitude that is important; it does not mat-
ter exactly what volume of information is produced daily, but it does matter, for
practical purposes, that it is very large, and getting much larger very rapidly. Nor
does it matter, for our purposes, whether the optimal group size for information
interaction is exactly Dunbar’s 150; though it does matter that it is about 150 rather
than the suggested alternative values of 30 or 500.
We may do better to forget numerological relations, and think of qualitative
patterns, with the numbers acting as a kind of aide-memoire: “statistical regu-
larities, observed in a context where social inuences play an important role”, as
Rousseau (2010, p. 2747) puts it. Or we may take the numbers as a clue, or introduc-
tion, to new theoretical perspectives, in the same way that Milgram’s small world
of 6 connections opens the way to the much wider idea of scale-free networks
following power laws (Mitchell, 2009).
In their LIDA2014 presentations, both Blaise Cronin and David Bawden cited
a quotation about the limitations of metrics. His quotation was Albert Einstein’s
remark that “not everything that can be counted counts, and not everything that
counts can be counted,” while David’s mentioned Václav Havel’s recommenda-
tion that we should have “a humble reverence for everything that we shall never
measure”. They amount to the same thing. Numbers will never tell the whole story,
in information or in any other context. But that should not prevent us from contin-
uing to seek for numbers, magic or otherwise, which capture the structures and
patterns of the information world.
Cited References
Åström, F. & Hansson, J. (2012). How implementation of bibliometric practice aects the role of
academic libraries. Journal of librarianship and information Science,45(4), 316–322.
Bawden, D. (2002). The three worlds of health information. Journal of Information Science,
28(1), 51–62.
Bawden, D., & Robinson, L. (2012A). Informetrics. In D. Bawden and L. Robinson, Introduction
to information science (pp. 163–185). London: Facet.
Bawden, D., & Robinson, L. (2012B). Basic concepts of information science. In D. Bawden and
L. Robinson, Introduction to information science (pp.63–89). London: Facet.
Information’s Magic Numbers: The Numerology of Information Science | 193
Bawden, D. & Robinson, L. (2013). “Deep down things”: in what ways is information physical,
and why does it matter for LIS?. Information Research,18(3), paper C03. Retrieved from
Beinhocker, E. D. (2006). The origin of wealth: evolution, complexity, and the radical remaking
of economics. Cambridge, MA: Harvard Business School Press.
Bhavnani, S. K., & Peck, F.A (2010). Scatter matters: Regularities and implications for the scat-
ter of healthcare information on the Web. Journal of the American Society for Information
Science and Technology,61(4), 659–676.
Blair, L. (1976). Rhythms of Vision, St. Albans: Paladin.
Bickart, K. C., Wright, C.I., Dauto, R.J., Bradford, C. D., & Feldman Barratt, L. (2011). Amygdala
volume and social network size in humans. Nature Neuroscience,14(2), 163–164.
Björneborn, L. (2006). “Mini small worlds” of shortest link paths crossing domain boundaries
in an academic Web space. Scientometrics, 68(3), 395–414.
Björneborn, L. & Ingwersen, P. (2001). Perspectives of webometrics. Scientometrics,50(1),
Borges, J. L. (1998), The Library of Babel. In J.L. Borges, Collected ctions, London: Allen Lane,
pp. 112–118.
Briscoe, B., Odlyzko, A., & Tilly, B (2006). Metcalfe’s Law is wrong. IEEE spectrum, July 2006,
pp. 26–31. Retrieved from
Broughton, V. (2006). The need for a faceted classication as the basis of all methods of infor-
mation retrieval. Aslib Proceedings,58(1/2), 49–72.
Caldarelli, G. & Catanzaro, M. (2012). Networks: a very short introduction. Oxford: Oxford Uni-
versity Press.
Case, D. O. (2 012). Looking for information: a survey of research on information seeking, needs
and behavior (3rd ed.),Bingley:Emerald.
Champernowne, D. G. (1933). The construction of decimals normal in the scale of ten. Journal of
the London Mathematical Society,8(4), 254–260.
Corby, K. (2003), Constructing core journal lists: mixing science and alchemy. portal: Libraries
and the academy,39(2), 207–217.
Cowan, N. (2001). The magical number 4 in short-term memory: a reconsideration of mental
storage capacity. Behavioural and Brain Sciences,24(1), 87–114.
Cronin, B. (1998). New Age numerology: a gloss on Apostol. Science Tribune, June 1998. Re-
trieved 1 July 2014 from
Cronin, B. (2000). Bibliometrics and beyond: some thoughts on webometrics and influmetrics.
Biomed central. Retrieved 1 July 2014 from
Cronin, B., & Sugimoto, C. R. (Eds.). (2015). Scholarly metrics under the microscope: From cita-
tion analysis to academic auditing. Medford, NJ: Information Today, INC/ASIST, pp. 976.
Davies, P. & Gregersen, N. H. (2010). Information and the nature of reality: from physics to
metaphysics. Cambridge: Cambridge University Press.
Davis, C. H. & Shaw, D. (eds.) (2011). Introduction to information science and technology.Med-
ford NJ: Information Today.
de Ruiter, J., Weston, G., & Lyon, S.M. (2011). Dunbar’s Number: group size and brain physiol-
ogy in humans re-examined. American Anthropologist,113(4), 557–568.
194 | David Bawden and Lyn Robinson
Dennie, D. & Cuccia, L.A. (2014). “Six degrees of separation”—Revealing a “small-world phe-
nomenon” through a chemistry literature search activity. Journal of Chemical Education,
91(4), 546–549.
Deutsch, D. (2011). The beginnings of innity: explanations that transform the world.London:
Allen Lane.
Dodds, P. S., Muhamad, R. & Watts, D. J. (2003). An experimental study of search in global so-
cial networks. Science,301(5634), 827–829.
Dunbar, R. I.M. (1993). Coevolution of neocortex size, group size and language in humans.
Behavioral and Brain Sciences,16(4), 681–684.
Dunbar, R. I.M. (2008), Mind the gap: or why humans aren’t just great apes. Proceedings of the
British Academy, vol. 154 (2007 lectures), 403–423.
Dunbar, R. I.M. (2012), Social cognition on the Internet: testing constraints on social network
size. Philosophical transactions of the Royal Society B,367(1599), 2192–2201.
Dunbar, R. I.M., Duncan, N. & Nettle, D. (1995), Size and structure of freely forming conversa-
tional groups. Human Nature,6(1), 67–78.
Egghe, L. (2005). Power laws in the information production process: Lotkaian informetrics.
Amsterdam: Elsevier.
Floridi, L. (2014). The 4th revolution: how the infosphere is reshaping human reality.Oxford:
Oxford University Press.
Ganz, J. & Reinsel, D. (2011). Extracting value from chaos. Retrieved 30 June 2014 from
Gleick, J. (2011). The information: a history, a theory, a flood. London: Fourth Estate.
Gould, M. (nd). Ramsey theory. Retrieved 12 September 2014 from
Granovetter, M. S. (1995). Getting a job: a study of contacts and careers (2nd ed.).ChicagoIL:
University of Chicago Press.
Grossman, J. (2014). The Erdös number project. Oakland University. Retrieved 12 September
2014 from
Guare, J. (1990). Six degrees of separation. New York NY: Vintage.
Hamilton, M. J., Milne, B.T., Walker, R.S., Burger, O., & Brown, J.H. (2007). The complex struc-
ture of hunter-gatherer social networks. Proceedings of the Royal Society B,274(1622),
Hedden, H. (2010). The accidental taxonomist. Medford NJ: Information Today.
Hilbert, M. (2014). What is the content of the world’s technologically mediated information
and communication capacity: how much text, image, audio and video? The Information
Society,30(2), 127–143.
Hilbert, M. & Lopez, P. (2011). The world’s technological capacity to store, communicate and
compute information. Science, 332(6025), 60–65.
Hill, R. A. and Dunbar, R.I. M. (2003). Social network size in humans. Human Nature,14(1),
Hjørland, B. (2007). Information: objective of subjective/situational? Journal of the American
Society for Information Science and Technology,58(10), 1448–1456.
Homan, K. & Doucette, L. (2012), A review of citation analysis methodologies for collection
management. College and Research Libraries,73(4), 321–335.
Huotari, M-L. & Chatman, E. (2001), Using everyday life information seeking to explain organi-
zational behaviour, Library and Information Science Research,23(4), 351–366.
Information’s Magic Numbers: The Numerology of Information Science | 195
James, K. (2006), Six degrees of information seeking: Stanley Milgram and the small world of
the library, Journal of Academic Librarianship,32(5), 527–532.
Jinha, A. (2010). Article 50 Million: an estimate of the number of scholarly articles in existence.
Learned Publishing,23(3), 258–263.
Kanai, R., Bahrami, B., Roylance, R., & Rees, G. (2012). Online social network size is reflected in
human brain structure. Proceedings of the Royal Society B,279(1732), 1327–1334.
Karinthy, F. (1929). Láncszemek (Chains). In F. Karinthy. Minden masképpen van (Every-
thing is dierent). Budapest: Atheneum Press. Reprinted in English in M. E.J Newman,
A. Barabási & D. J. Watts. The structure and dynamics of networks. Princeton NJ: Princeton
University Press, 2006, pp. 21–26.
Kauman, S. (1996), At home in the universe: the search for laws of self-organization and com-
plexity. Oxford: Oxford University Press.
Kauman, S. (2010), Reinventing the sacred: a new view of science, reason and religion, New
York NY: Basic Books.
Kilmister, C. W. (2005). Eddington’s search for a fundamental theory: a key to the universe.
Cambridge: Cambridge University Press.
Lee, N. H. (1969). The search for an abortionist, Chicago IL: University of Chicago Press.
McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: homophily in social
networks. Annual Review of Sociology,27, 415–444.
Milgram, S. (1967). The small world problem. Psychology Today,1(1), 61–67.
Miller, G. A. (1956). The magical number 7 plus or minus 2: some limits on our capacity for pro-
cessing information. Psychological Review,63(2), 81–97.
Mitchell, M. (2009). Complexity: a guided tour. New York NY: Oxford University Press.
Nicolaisen, J. & Hjørland, B. (2007). Practical potentials of Bradford’s Law: a critical examina-
tion of the received view. Journal of Documentation,63(3), 359–377.
Pfender, F., & Ziegler, G. M. (2004). Kissing numbers, sphere packings and some unexpected
proofs. Notices of the American Mathematical Society,51(8), 873–883.
Pickover, C. A. (2012). The math book. New York NY: Sterling.
Powell, J., Lewis, P. A., Roberts, N., García-Fiñana, M., & Dunbar, R. I.M. (2012). Orbital pre-
frontal cortex volume predicts social network size: an imaging study of individual dier-
ences in humans. Proceedings of the Royal Society B,279(1736), 2157–2162.
Priego, E. (2012). Insights from “The Numbers Game”, lecture by Blaise Cronin. Altmetric
blog. Retrieved 1 July 2014 from
Roberts, S. B.G., Dunbar, R.I. M., Pollet, T., & Ruppens, T. (2009). Exploring variations in active
network size: constraints and ego characteristics. Social Networks,31(2), 138–146.
Rousseau, R. (2010). Informetric laws, Encyclopedia of library and information sciences
(3rd ed.). London: Taylor and Francis, 1:1, 2747–2754.
Savolainen, R. (2009), Small world and information grounds as contexts of information seeking
and sharing. Library and Information Science Research,31(1), 38–45.
Schnettler, S. (2009). A structured overview of 50 years of small-world research. Social Net-
works,31(3), 165–178.
Schultz-Jones, B. (2009). Examining information behaviour through social networks: an inter-
disciplinary review, Journal of Documentation,65(4), 592–631.
Shirky, C. (2003). A group is its own worst enemy. Clay Shirky’s writings about the Internet.
Retrieved 1 July 2014 from
196 | David Bawden and Lyn Robinson
Six degrees of separation. (n.d). In Wikipedia. Retrieved 29 June 2014 from http://en.wikipedia.
SIMS (2003). How much information? School of Information Management and Systems, Univer-
sity of California Berkeley. Retrieved June 30 2014 from
Stock, W.G., & Stock, M. (2013). Handbook of information science. Berlin: Walter de Gruyter.
Tegmark, M. (2014). Our mathematical universe, London: Allen Lane.
Tetlow, P. (2007). The Web’s awake: an introduction to the eld of web science and the concept
of web life. Hoboken NJ: Wiley.
van Mierlo, T. (2014). The 1% rule in four digital health social networks: an observational study.
Journal of Medical Internet Research,16(2), e33. Retrieved 30 June 2014 from http://www.
von Baeyer, H. C. (2003). Information: the new language of science. London: Weidenfeld &
Watts, D. J., & Strogatz, S. H. (1998), Collective dynamics of ‘small world’ networks. Nature,
393(6684), 440–442.
Wiesstein, E. W. (n.d.). Kissing number. Mathworld. Retrieved 1 July 2014 from http://
Wellman, B. (2012). Is Dunbar’s number up?. British Journal of Psychology,103(2), 174–176.
Zhou, W-X., Sornette, D., Hill, R.A., & Dunbar, R.I.M. (2005). Discrete hierarchical organization
of social group sizes. Proceedings of the Royal Society B,272(1561), 439–444.
Zhu, D., Wang, D., Hassan, S., & Haddawy, P. (2013). Small-world phenomenon of keywords
network based on complex network. Scientometrics,97(2), 435–442.
... It may be altmetrics, considering non-publications (data files, videos, presentation, event, performance, interviews, etc.), or considering publications that do not appear in citation databases (scholarly monographs, textbooks, commissioned research reports, newspaper articles, etc.). Despite warnings about descending into a "new age of numerology," (Bawden and Robinson, 2016) uses are made of collapsed scores, where several indicators are combined in one number because it is useful [2]. ...
Purpose The purpose of this paper is to acknowledge that there are bibliometric differences between Social Sciences and Humanities (SSH) vs Science, Technology, Engineering and Mathematics (STEM). It is not so that either SSH or STEM has the right way of doing research or working as a scholarly community. Accordingly, research evaluation is not done properly in one framework based on either a method from SSH or STEM. However, performing research evaluation in two separate frameworks also has disadvantages. One way of scholarly practice may be favored unintentionally in evaluations and in research profiling, which is necessary for job and grant applications. Design/methodology/approach In the case study, the authors propose a tool where it may be possible, on one hand, to evaluate across disciplines and on the other hand to keep the multifaceted perspective on the disciplines. Case data describe professors at an SSH and a STEM department at Aalborg University. Ten partial indicators are compiled to build a performance web – a multidimensional description – and a one-dimensional ranking of professors at the two departments. The partial indicators are selected in a way that they should cover a broad variety of scholarly practice and differences in data availability. Findings A tool which can be used both for a one-dimensional ranking of researchers and for a multidimensional description is described in the paper. Research limitations/implications Limitations of the study are that panel-based evaluation is left out and that the number of partial indicators is set to 10. Originality/value The paper describes a new tool that may be an inspiration for practitioners in research analytics.
Full-text available
This paper examines social network size in contemporary Western society based on the exchange of Christmas cards. Maximum network size averaged 153.5 individuals, with a mean network size of 124.9 for those individuals explicitly contacted; these values are remarkably close to the group size of 150 predicted for humans on the basis of the size of their neocortex. Age, household type, and the relationship to the individual influence network structure, although the proportion of kin remained relatively constant at around 21%. Frequency of contact between network members was primarily determined by two classes of variable: passive factors (distance, work colleague, overseas) and active factors (emotional closeness, genetic relatedness). Controlling for the influence of passive factors on contact rates allowed the hierarchical structure of human social groups to be delimited. These findings suggest that there may be cognitive constraints on network size.
Full-text available
This lecture presents the text of the speech about humans and apes delivered by the author at the 2007 Joint British Academy/British Psychological Society Annual Lecture held at the British Academy. It comments on the claim that an evolutionary perspective is not a competing paradigm for conventional explanations in the social sciences, and explains the why humans are so different from other apes and monkeys, despite the fact that we share so much of our evolutionary history with them.
Dealing with information is one of the vital skills in the 21st century. It takes a fair degree of information savvy to create, represent and supply information as well as to search for and retrieve relevant knowledge. How does information (documents, pieces of knowledge) have to be organized in order to be retrievable? What role does metadata play? What are search engines on the Web, or in corporate intranets, and how do they work? How must one deal with natural language processing and tools of knowledge organization, such as thesauri, classification systems, and ontologies? How useful is social tagging? How valuable are intellectually created abstracts and automatically prepared extracts? Which empirical methods allow for user research and which for the evaluation of information systems? This Handbook is a basic work of information science, providing a comprehensive overview of the current state of information retrieval and knowledge representation. It addresses readers from all professions and scientific disciplines, but particularly scholars, practitioners and students of Information Science, Library Science, Computer Science, Information Management, and Knowledge Management. This Handbook is a suitable reference work for Public and Academic Libraries.
Many scientists regard mass and energy as the primary currency of nature. In recent years, however, the concept of information has gained importance. Why? In this book, eminent scientists, philosophers and theologians chart various aspects of information, from quantum information to biological and digital information, in order to understand how nature works. Beginning with a historical treatment of the topic, the book also examines physical and biological approaches to information, and its philosophical, theological and ethical implications.
The central thesis of The Web's Awake is that the phenomenal growth and complexity of the web is beginning to outstrip our capability to control it directly. Many have worked on the concept of emergent properties within highly complex systems, concentrating heavily on the underlying mechanics concerned. Few, however, have studied the fundamentals involved from a sociotechnical perspective. In short, the virtual anatomy of the Web remains relatively uninvestigated. The Web's Awake attempts to seriously explore this gap, citing a number of provocative, yet objective, similarities from studies relating to both real world and digital systems. It presents a collage of interlinked facts, assertions, and coincidences, which boldly point to a Web with powerful potential for life. © 2007 the Institute of Electrical and Electronics Engineers, Inc.
While there is a considerable body of literature that presents the results of citation analysis studies, most researchers do not provide enough detail in their methodology to reproduce the study, nor do they provide rationale for methodological decisions. In this paper, we review the methodologies used in 34 recent articles that present a "user study" citation analysis with a goal of informing collection management. We describe major themes and outliers in the methodologies and discuss factors that require careful thought and analysis. We also provide a guide to considerations for citation analysis studies, so that researchers can make informed decisions.
Who are we, and how do we relate to each other? This book argues that the explosive developments in Information and Communication Technologies (ICTs) is changing the answer to these fundamental human questions. As the boundaries between life online and offline break down, and we become seamlessly connected to each other and surrounded by smart, responsive objects, we are all becoming integrated into an "infosphere". Personas we adopt in social media, for example, feed into our 'real' lives so that we begin to live, as Floridi puts in, "onlife". Following those led by Copernicus, Darwin, and Freud, this metaphysical shift represents nothing less than a fourth revolution. "Onlife" defines more and more of our daily activity - the way we shop, work, learn, care for our health, entertain ourselves, conduct our relationships; the way we interact with the worlds of law, finance, and politics; even the way we conduct war. In every department of life, ICTs have become environmental forces which are creating and transforming our realities. How can we ensure that we shall reap their benefits? What are the implicit risks? Are our technologies going to enable and empower us, or constrain us? This volume argues that we must expand our ecological and ethical approach to cover both natural and man-made realities, putting the 'e' in an environmentalism that can deal successfully with the new challenges posed by our digital technologies and information society.