COHERENT INFORMATIONAL ENERGY AND ENTROPY
| Date | 01 April 1980 |
| Pages | 293-312 |
| Published date | 01 April 1980 |
| DOI | https://doi.org/10.1108/eb026700 |
| Author | AUREL AVRAMESCU |
| Subject Matter | Information & knowledge management,Library & information science |
COHERENT INFORMATIONAL ENERGY AND
ENTROPY
AUREL AVRAMESCU
Bucharest—Otopeni
Aiming to provide a common theoretical foundation for all known biblio-
metric
laws,
the author starts from a systemic view of the information transfer
process and assimilates it with a physical diffusion process, in particular the
conduction of heat in solids. Previous literature induces in the properly
ranked
space
of new authors an interest potential (temperature) confirmed by
exchange of reference-citation pairs, and driving a controlled information
flow. The model gives its distribution for given initial and borderline con-
ditions,
allowing at the
same
time the establishment of new definitions for in-
formational energy and entropy, which are coherent with the corresponding
physical ones. The theory shows that impulsory energy supply is bound to
negative entropy inflow which brakes the normal entropy 'production' in the
'dissipative' structures of the considered system. In this way the introduction
of information into concrete thermodynamic systems analysis can hopefully
be expected.
INTRODUCTION
WHILE IN THE last four decades excellent scholars and librarians have been
able to discover many a statistical well-proven bibliometric law, such as Lotka's1
and Bradford's,2 researchers in this field were soon aware of the need to clarify
matters, i.e. to find a common theoretical basis for all bibliometric relations. A
first result of their widespread effort was the assimilation of well-known distribu-
tions,3
such as Zipf's in the case of the main Bradford law.4 In order to obtain a
wider foundation, the author forwarded the idea of assimilating the dissemina-
tion of scientific information with a physical diffusion process,5 a tentative pro-
posal which consequently proved to be highly successful.6 I am obliged to men-
tion here as a counterpart the remarkable contribution of Derek Price, who
arrived at such a foundation by quite different ways.7
The advantage of assimilating scientific information dissemination with the
diffusion process resides in the huge power of the theory of heat conduction
coined by J. B. Fourier in 1822 and developed since then by many outstanding
scholars, especially in deducing a large number of solutions to various implied
boundary value problems. One famous example remains the design of the first
transoceanic telegraph cable by Lord Kelvin in 1855.8
Having in mind the intention to present the main relations of the assimilated
information diffusion theory, the author will try to keep the main text of this
paper within the limits which could be understood by experienced librarians.
Therefore mathematical deductions which complete the text will be transferred
to Appendices.
The structure and sequence of parts delineated will follow the main path of any
scientific research: organized observations, statistical surveys, empiric relations
Journal
of
Documentation,
Vol. 36, No. 4, December
1980,
pp. 293-312.
293
JOURNAL OF DOCUMENTATION Vol.
36,
no.
4
and laws, search for analogue or similar processes based on a verified theory,
assimilation with well-fitting models which offer corresponding solutions, inter-
pretation and practical application of the results.
In this frame we take the opportunity to re-open previous discussions on
analogies and differences between the Clausius-Boltzmann type thermodynamics
and C. Shannon's informational entropy, recommending
a
new definition which
proves to be physically coherent and can put an end to older and more recent
confusion.
The new definitions of informational energy and entropy entitle us to hope
that a joint treatment of information and thermodynamic or even economic
systems will soon be made possible. This hope is supported by the recent sub-
stantial progress made in non-equilibrium thermodynamics forwarded by the
remarkable research groups led by Ilya Prigogine (Nobel Prize 1977) at the Free
University of Brussels,9 and the research concerning the inference of the entropy
law in economic processes conducted by N. Georgescu-Roegen at Vanderbild
University in Nashville, Tennessee, USA.10
AVAILABLE STATISTICAL DATA
It is to the merit of experienced librarians that they have carefully ordered and
classified publications and thereby observed recurrent relations between
numerous parameters emerging in bibliographic activities. Even when these
parameters have a discrete character, their variations can be expressed and
studied by the theory of
distributions3
and that of probabilities. In this way we
recognized that the number of authors,
books,
journals, papers etc. does indeed
exponentially increase, but with different doubling times (or time constants). So,
by comparing distributions we are aware that the increasing rate of scientific
papers is 50% higher than that of journals.
Bradford's law of scattering appeared as a result of laborious statistical surveys
and despite its ambiguous formulation initiated a great number of theoretical
studies and even criticism, culminating in E. A. Wilkinson's memorable call:
'What is Bradford's law?'11 Nevertheless, since its issue the law has been
frequently tested and found consistently sound, except in the core and the final
phase, where strange deviations have appeared.12 But, even including Naranan's
'interpretation',13 nobody could 'explain' why it is valid. This question
is
equally
justified as in the case of Kepler's empirical laws, which have been 'explained' by
Newton's fundamental law of universal gravitation.
Having this in mind, the author was encouraged to find a model and a theory
for the diffusion of scientific work between readers and authors of new papers.
Soon after the creation of the Institute for Scientific Information (ISI) in Phila-
delphia, the
Science Citation
Index and the computer memory used by its editors
began supplying
a
wealth of statistical data concerning authors, papers, references,
citations etc. revealing many a correlation between previous and new scientific
work. For all these activities the word
Bibliometrics
was coined. Among other
advantages, these data allow one to objectively illustrate the progress of science
in its main disciplines and as a whole. The generalization of such research activi-
ties is denoted today as
Scientometrics,
the name also of
a
recently issued inter-
national journal edited by Elsevier and the Publishing House of the Hungarian
Academy of Sciences.
Price presented a remarkable paper exploiting such a survey, on the rate of
294
To continue reading
Request your trial