'Science communication' as defined here is formal communication within the academia, not communicating science to (or with) the public at large. We are witnessing a transitional period for science communication due to the impact of the information technology [1] . While information technology may be transforming communication techniques, formal communication remains critical to the advancement of science. To quote the most recent source, an international encyclopaedia [2] :

"Scientometrics can be defined as the study of the quantitative aspects of scientific communication, R&D practices, and science and technology (S&T) policies." Hence, the communication within science is one of the three aspects of science studies in a quantitative manner. While the communication may be manifold (from informal, personal, to purely formal), the formal one through scientific journals will be discussed here.

The formal scientific communication structure is shaped by secondary information services - the databases. There are no clear selection criteria statements when it comes to selecting for secondary information the journals published within peripheral scientific communities.  

The term “periphery” has no simple definition in the context of science. It was used earlier [4] with a very strong connotation referring to less developed countries. But the difference between the "centre(s)" and the "periphery" was stressed by invoking the element of size [5]: " Although the concept of periphery is usually associated with the Third World countries that are relative latecomers to western science, many small, economically advanced countries (in Europe for instance) for structural and cultural (or linguistic) reasons are also in a peripheral position.” In many respects both types of countries face similar problems with respect to the automatic adoption of evaluation procedures developed in the centre for assessing its scientific activity.

Here is, however, the description of the term in a more qualitative (and perhaps substantial) way [6] . There are, at least, three characteristics of scientific periphery:

•  (I) its smallness,

•  (II) lack of societal equilibrium, and

•  (III) communication barriers.

(I) ( Smallness ) The scientific community is relatively small in regard to the fields of current research. The sheer size of the country (geographic and/or demographic), is not of sole importance, but rather the structure of its scientific endeavour. Peripheral scientific communities have a sub-critical research mass in many of the research fields pursued. Under such circumstances there is frequently a lack of qualified scientists to take part in the peer review process. Instead, subjective "in-person assessments" [7] come to the forefront.

(II) ( Societal non-equilibrium ) Owing to their smallness peripheral scientific communities lack self-regulatory mechanisms which are otherwise common in well developed modern societies where the science sphere is one of the social factors within the decision making procedures.

(III) ( Communication barriers ) Within the human civilization's time scale science has become a lasting and dynamic world process of cognition. Any barrier against bringing closer the periphery to the core of the science process hampers the very substance of this human endeavour. The ideological and political barriers may be regarded as special cases, and they may or may not persist for long, but the sociocultural barriers are very serious, and among them the language could be regarded as the most important: "…to replace the indigenous language by a foreign one is dangerous in science because it hampers the development of the all-embracing modern culture within the ethnic group…(and)…to make negligible the cultural barriers preventing the social influence of science, it (science) must be integrated within the culture(s) of particular countries…" [8]

Can the literature of the science periphery be integrated into the science communication fabric? Positively articulated replies to this question have come from the most advanced part of the world [9 , 10] , but the cultural, technical and financial hurdles are serious. An impressive overview of the real-life difficulties in peripheral science communities is presented in [11] . In [12 ] it is stated that "Almost 50 percent of all African research reports are published in local scientific journals that are not listed in information data bases."

Introduction

It would seem desirable to determine if there are any journals evaluation/ranking studies outside the databases management systems, which could be used in constructing a common methodology for improving the databases journal selection procedures. In pursuing this aim there is hope that if a common methodology for journals evaluation and ranking could be discerned it might perhaps help in ameliorating the status quo .

The rationale for this review is twofold. First, externally (with respect to scholarly journals origins), the science communication structure concerning the peripheral scientific communities cannot be neglected within the science studies in general; second, internally , from the practical point of view, most of the journals in peripheral scientific communities are directly financed by public money (as opposed to the well-developed world). There is thus a need for an efficient and, as far as feasible, an objective method as a basis for indigenous financial decision making in support of journals (and/or for their inclusion in international databases as mentioned here

initially).

The last decade or so has witnessed several attempts, from the less developed countries, to devise methods for evaluating the indigenous journals. Prominent among these is the Latin American case [13] .

Journals are products of cooperation within the "triangle" of -authors-editors (+ publishers)-referees-. For the present purpose the editing-and-publishing ( EP ) process will be treated as far as it is demonstrated solely through the physical appearance of the journals.

Evaluation Indeterminacy

About quarter of a century ago Francis Narin pointed out an inherent indeterminacy in the evaluation of “scientific advance”. The paper was published in the very first issue of Scientometrics [16] , nowadays considered a leading journal for quantitative science studies.

There is no mathematical expression of Narin's indeterminacy, so that the concept can only be described qualitatively (more detailed in [16 , 17 , 18 , 19]) . Narin depicts it in a two-dimensional graph of the “More relevant (methods) to true measurements of scientific advances” (the ordinate) and the “More objective of the methods used in the assessment” (the abscissa).

In the graph, a downward “curve” begins at the lowest objectivity with concurrently high relevance of methods ascribed mainly to expert opinions. The relationship (“curve”) between the two variables descends further on in relevance, at first slowly, while the objectivity increases. The concluding part of the curve is within a rather narrow range of high objectivity, while the relevance diminishes sharply to its lowest “value” (at the highest objectivity), ascribed to methods based on quantifiable parameters (simple counts).

The whole “interdependence” is substantiated by Narin's intuitive choice and sequencing of various methods in assessing the “scientific advance”. The beginning of the “curve” is depicted by methods exploiting several shades of interviews and surveys, what we shall call here the questionnaire methods ("q") . In his papers Michael Moravcsik [20 , 21 , 9] calls the indicators derived by these methods as perceptual , which “implies a personal evaluation, by inspection, on the part of knowledgeable investigators of the particular situation to be assessed. This kind of assessment is often referred to as ‘peer review'” (p.172 in [20] ). The maximal objectivity at the lowest relevance end is substantiated by various methods of quantifying the scientific productivity. These methods are called databased type, and, sometimes, “quite misleadingly, objective, including bibliometric measure, patent counts, production statistics, counts of literate persons, etc.”. (p.172 in [20] ).

As journals are the primary public record of the science process dynamics, Narin's indeterminacy may be expected (tough its generality not yet tested) to hold in journal evaluation as well.

Let us turn now briefly to a puzzling question, which was not addressed in [16] . How come at all such an inverse relationship between the objectivity of a method and its relevance to the determination of the value of a (societal!) manifestation like science (and maybe in case of scholarly journals as well)? One can rationalise, perhaps, that the objectivity increases with the simplification of the observable manifestation. It occurs, however, at the expense of the relevance, because it diminishes simultaneously as only its outer, ever more simplified, manifestation is being observed. The comprehending of the underlying “heart of the matter” is thus gradually being lost.

Journals are emanating from and for the science endeavour. Their role as constituents of formal science communication channels is only part of the science story, though an important one. So, what "property" can one invoke to evaluate a journal w i t h i n scientific communication channels? We choose here - the journal's scientific communicability.

In the case of journal evaluation this inverse relationship (between the journal's scientific communicability and the objectivity of its assessment) seems to break into two distinct sections [17] . As in the case of Narin's original paper, we intuitively select the available approaches to evaluate the scientific communicability of journals. Narin dealt with a broad definition of "scientific advance". Journals are a reflection of scientific advance if "the whole sample" is taken into account. In such a case (of all scholarly journals) one would expect one monotonous "curve" depicting the indeterminacy relationship like in the case of "scientific advance".

However, when intuitively selecting one particular evaluation "method" the researcher simultaneously and unwillingly selects subsamples of journals, which may lead to discontinuity in the indeterminacy relationship as it was observed here with the methods for evaluating scientific communicability of journals. This may account for a dichotomy in applying Narin's indeterminacy when evaluating scholarly journals.

One approach, that of high objectivity/low relevance, does comply with the inverse relationship. The other, however, “defies” such an uncertainty in that for the starting region at rather low objectivity both the relevance (for assessing the scientific communicability) and the (methodological) objectivity increase within a rather narrow range of the latter. This framework, then, yields roughly two juxtaposed classes of methods for evaluating journals: C lass A of low objectivity, and C lass B of higher objectivity.

The Methods

Class A methods - low objectivity

Four levels in a concurrent increase of both objectivity and relevance could be discerned among the methods “defying” Narin's indeterminacy as follows [17] :

(i) Journal selections by the local public fund sponsoring agencies (through their "expert committees"); this is the case of creating perceptual indicators in Moravcsik's sense.

Moravcsik's wording (p. 174, in [20] ) adds weight to our putting the indigenous “expert committees” journal ranking at the lowest level of both the relevance and the objectivity: “Perceptual indicators, that is peer reviewing, are also on shaky grounds in the context of developing countries. Scientific communities in individual countries are mostly too small to allow internal peer reviewing … developing countries are often reluctant to turn to external peer reviewing, because those in charge of science policy are not acquainted sufficiently with the world-wide scientific community to know whom to ask, because going outside the country is thought to reflect on the national sovereignty, because there is fear that the request for outside help will be rebuffed, and because financial resources may not be available to organise such external peer reviews. In some cases external review teams arranged through international organisations turned out to be both ignorant of and insensitive to the local conditions under which scientific work must be performed in the developing countries.”

(ii) By domestic (national) bibliographical and indexing/abstracting services with a higher or lower relevance according to their expected readership - international, or domestic, respectively.

The national bibliographic services (if there are such) accumulate the experience in exercising good judgment to define the pool of domestic journals for their bibliographic processing. Such approaches yield the widest possible journal selection. If bibliographic or indexing services exist which are meant for the readership abroad, the relevance as to scientific communicability may be expected to be somewhat higher than in case for domestic readership. In either case the objectivity is of the same degree and higher than in (i) because those making a certain choice of journals are not individually under pressure from interest groups.

(iii) By international abstracting or bibliographic services (databases); the objectivity and the relevance in assessing the degree of the journals' scientific communicability by the database managements selections may be taken as being higher than in the two preceding cases because the assessment is further removed from the domestic publishing scene. Such selections lead to very restricted lists of domestic journals.

In accepting this evaluation approach, without knowing the database's criteria for indexing , the domestic science policy flounders because many journal titles are omitted from the international scene. This has become commonplace in the pertinent professional literature, such as [23 , 26 , 27 and 13] .

(iv) Inclusion of journals by the Institute for Scientific Information, ISI, (Philadelphia, USA) in their citation indexes, attains the highest relevance within this rather low objectivity range ( Class A methods ) yielding the most stringent set of domestic (peripheral) journals, frequently only a handful, if any.

This kind of journal selection results in most distorted science policy decision-makings in science periphery. The problem is addressed in [28] (pp. 58-61) with the conclusion: “All these recent findings substantiate the thesis that the bibliometric indicators, especially the SCI , do not accurately assess the scientific output from the periphery, especially from the DC's and that local science, far from being synonym of poor science, (is) at least as important as international science in the context of a developing country, and should thus be taken into account.” Several similar very succinct statements with respect to science policy can be found in [26] .

The existence of the ISI journal pool selection has been analysed until recently only from an academic point of view, especially in relation to the so called “impact factors” (IF) of journals as derived from the ISI citation data. There are warnings against making indiscriminte use of IF, but all those "academic discussions" did not uncover IF's economic (and cultural) consequences of journal publishing. The (bad) economic side of the IF- (i.e. ISI's together with some largest publishers') role is a "frozen", not anymore a free journal market. This was convincingly shown only during the past couple of years. [29]

The economic consequences aside, there appears to be (for evaluation purposes) a rather simple way out of this “citation deadlock” [30] : the proposal is to set up a regular annual ‘Extra- CI ' directory, which would record titles of those journals, which have been cited through the regular CI -journals pool, although they themselves are not taken-up for regular processing by the ISI . Such annual indexes would also be useful to various secondary information services in deciding on their journal coverage.

The ongoing “journals citation equilibrium,” as presented through the long publishing activity of the Institute for Scientific Information ( ISI ), may, therefore, be beneficial for a purpose for which the citation indexes were not intended in the first place - the science policy. Namely, the most stringent journal lists could be obtained by the science process itself (i.e., with its referencing/citations practices), alleviating thus unwanted distortion effects in science policy within peripheral scientific communities.

When (or if) such state of affairs materialises, i.e., when the “Extra- CI ” journal (annual) indexes appear, journal selection through them would be of highest attainable objectivity and relevance.

Until then, domestic journals can only be categorised crudely as to their (perceived) scientific communicability by the available selections of journals from scientific periphery, i.e. by the Class A methods (i) to (iv) above. Methods (i) and (ii) will yield the most abundant set, whereas the most stringent subset is obtained according to (iii) and (iv) .

Class B methods - higher objectivity

Let us turn now to the other group of methods, the Class B methods “obeying” Narin's indeterminacy. Here, the highest objectivity (of least relevance) is ascribed to the journals' e diting and p ublishing characteristics ( EP) - discerned de visu from the journals themselves.

In saying this we assume that the published appearance of a journal reflects the outcome of a multifaceted editing and publishing effort. To put it otherwise, the published journal reflects (as an external manifest) the degree of the journal's scientific communicability.

Of higher relevance for scientific communicability is the data derived from the l iterature r eferences ( LR) in the papers of a given journal. The highest attainable objectivity is attributed to these two methods - EP and LR - because there is no mediator whatsoever between the evaluator and the journal.

The difference in the relevance status of these two methods stems from the source reflected in the outer manifest: for EP it is the editing/publishing triangle within which the authors are involved only to the (hidden) extent they react to the referees' suggestions; otherwise, EP is independent of the authors. For LR the authors are in the forefront, because the selection of literature (references) in their papers is only theirs (with some “brushing-up” by the referees). Hence a higher relevance to the scientific process in LR than in EP .

The third method, the use of citation data ( CD ) , is of least objectivity but highest relevance in Class B method . There is an important distinction from the use of citation data in (iv) of Class A methods . Namely, the latter makes use merely of the source journal titles within the existing arbitrary selection by the database management. It is therefore at lower objectivity and relevance than is the case for the CD method in the Class B methods: here, the citations given to the individual journals are from an “unsolicited survey” within the citing space. The latter, however, is presently limited only to citations by the ISI- journal titles from the arbitrary ISI- selection. Citations from the ISI- source journals to those (mostly from peripheral scientific communities!) that are not ISI -source journals are completely neglected. The situation with respect to the latter could be remedied, as mentioned above (for a recent printed explanation see [31] ).

“The relevance” within the context of this review means the relative degree to which the data obtained from the journals appearance (i.e. their manifest, not their content) are indeed reflecting the journals' scientific aspect of their communicability. This may be expected to be judged more adequately by the scientists themselves, here the editors, referees and the authors. Such may be the case in the science centre(s). However, the societal "equilibrium" of the science process from the centre(s) is lacking within the periphery (see above (II) Societal non-equilibrium ). However, the highest attainable objectivity is necessary for cross-disciplinary and international comparisons, and that is possible only with the scicomm approach by confronting the observer directly with the journal.

The Comparisons

Is there a consensus about the method to be preferably used for evaluation of journals from science periphery? Not yet, but it is this author's view that the time is appropriate for defining such a method. In doing so we concentrate on the Class B methods - those of higher objectivity. They will be referred to hereinafter as the scicomm methods, indicating their potential value for estimating the sci entific comm unicability of the journals. The Class A methods of lower objectivity will not be analysed furthermore.

Various shades of the scicomm approach have been described in [32] , [33] and [34] . The first author of [32] presented their main findings also in [35] , where other pertinent references are given. In [34] there is reference to the earlier work of the same group. In all the three cases the direct evaluator-object approach has been used, with the questionnaire (“ q ”) tools applied partly, too, in [32] and [33].

The EP -indicator

Of all the EP -parameters found to be listed so far [14] , those four described here ( PR , L , R , FA) are considered to be necessary and sufficient in defining the EP-indicator for an ordinal scale. As said, one more parameter, however, may also have to be included. It is the presence (or absence) of periodic and/or cumulative content indexes as used in [32] and [33] .

If coverage by secondary information services is involved at the outset of evaluation it interferes with the application of the independent, indigenous methodology. Afterwards, one may (safely) compare the two rankings.

The (ordinal) ranking

Once the basic parameters constitutive of the EP indicator will have been agreed upon in pursuit of a common methodology the next step is to assign quantitative values to each of the parameters. This has already been done in the three reports discussed here, but only in an intuitive way. (This is a difficult problem that will have to be addressed by expert statisticians.)

The advantage of the ordinal scale (whatever the final choice of the attributed weights) is that distributions of evaluated journals can be shown graphically [34] as journals frequencies (the ordinate) along the ordinal scale of their scientific communicability (the abscissa). Thus stratification of journals according to their derived scientific communicability can be discerned.  

The LR -level parameters

When clearer stratification of journal groupings is necessary the other indicator - that of the literature referencing ( LR) structure - can be taken into account. This is particularly important because of the well-documented dichotomy in EP between the (natural) sciences journals and those of the social sciences and humanities [34] . It should be emphasized that the EP- and LR-indicators cannot be blended together directly, owing to their ordinal and digital scales respectively, but if necessary they can be used together judiciously.

We have thus entered the realm of the authors of scholarly writing. In comparison to the preceding analysis of the EP -level parameters and of the resulting EP-indicator , the references in scientific papers are most directly related to the scientific communication. As such these parameters and the indicator to be derived from them are of considerably higher relevance to estimating the scientific communicability of a journal. At the same time, objectivity remains the highest possible.

The LR -level in evaluating journals is used neither in [32] nor in [33] . It has been explored in [34] . From many a possible aspect (facets) of the references structure as described in [17] and [18], only two were retained in [34] .  

The referenceless papers

To count how many papers are without any references in a journal issue is certainly a simple and fast bibliometric technique. There is a widespread agreement that such papers can hardly be counted as scientific, and are therefore of little scientific communicability. This parameter can be calculated as percentage of referenceless papers from an initial number of papers (whatever the pool is) in a journal under scrutiny. The parameter is thus of a negative connotation in estimating scientific communicability of journals from peripheral scientific communities.

Papers above a threshold number of references

In [34] it was explained why a threshold of five references per paper was chosen. The choice will depend on a particular science periphery setting and the purpose of using that parameter, which is of a positive connotation for journal evaluation.

The LR-indicator & scale

Here, in distinction from the EP-indicator, there is neither a composite LR-indicator nor a ranking scale. The reason is that each of the LR- parameters stands for itself as numerical, not ordinal. Each of them may only be used in addition to the (composite) EP-indicator .

The CD- level parameters

Citation Data ( CD ) are mentioned here already within the description of the Class A and B methods . It was pointed out that the present status of the c itation i ndexes produced by ISI leaves much to be desired with respect to journals from the science periphery. It was mentioned that introducing a "value-added" Extra- CI directory would mean exploiting to the fullest the "science communication equilibrium" attained within the publications of the Institute for Scientific Information ( ISI ). In such a case the CD -level would become a joint between the Class A and Class B methods at a very satisfactory objectivity and the correspondingly highest possible relevance for estimating the worldwide scientific communicability of the domestic journals from science periphery. Until then, the CD -level proper, i.e. making use also of all the presently "invisible" (but recorded) space of citations given from the ISI -journals to those which are not in the ISI processing pool - requires financial and technical means usually not available to science communities at the periphery. A good example of how it could be done, though not at the science periphery, is described in [39]. The latter application was made possible by courtesy of the author of [39] for the final stage of evaluating the domestic journals in [34] .

The CD -level was not used in [33] with an explicit statement to the effect that no journal from that domestic pool had been incorporated for processing by ISI . Neither was this evaluative level dealt with in [32] .

On fusing the parameters or indicators

There are statistical methods which perhaps could be used in "fusing" the EP and LR parameters or the corresponding indicators, methods like the Correspondence Analysis and Adjusted Scores approach. However, no attempt is known to this reviewer for "fusing" properly the various parameters or indicators (described in section "The Comparisons") into a single indicator of scientific communicability. Recently, however, the possibility of using "fuzzy set" mathematical approach has been suggested [40]. In the Introduction the authors say: "Unfortunately, there is no consensus on how to best conduct the journal evaluation." Further, the dichotomy between the "subjective" and "objective" methods is emphasized, the "objective" covering, according to the authors - the citation data.

A reflective summing-up

In kind of a "last will" statement Moravcsik said: “Because of the decentralized nature of most of the proposals, it is difficult to assess to what extent these recommendations have been implemented in the last two years since the Philadelphia meeting [9] . It is possible that there has been more taking place than meets the eye. Nevertheless, judging from the visible evidence, it appears that hardly anything of the proposed programs has been converted into reality so far. The aim of publicizing this effort at this conference, therefore, is to give it renewed publicity in the hope that eventually some action will follow the words. Success depends entirely on personal initiatives." [41]

Now, 17 years after "Philadelphia" it seems appropriate to ask ourselves (from the science periphery, at least) "where are we heading to?" A positive stand towards evaluation (with the aim for subsequent betterment) of scientific communicability of journals from the science periphery has been adopted here and, hopefully, shown to be the choice based on many quoted publications. Though, quite a sizable literature has accumulated in the past couple of decades dealing with journals ranking from the science centres, too, usually by citation data. It appears that some independent evaluation is to be taken note of by the "gate keepers" within the science centre. However, within the science periphery a purely "evolutionary" way will take too much of precious time bearing in mind the dynamics of the world science process.

Much as it is true that "…success depends entirely on personal initiatives…" [41] some infrastructure must be there to accept and implement the initiatives. The example of Latin America and the Caribbean is far ahead in that respect [42 , 43 , 13] . Most notable is the "electronic visibility" achieved within a few recent years bearing witness to the initial statement in this review [1] . Two URLs are of particular interest, because one [44] shows the bibliographic advancement in recording printed publications, whereas the other [45] deals with the transformation of the journals into the electronic format, summarizing also the evaluation technique for selecting appropriate journals. According to information kindly provided (on 16 Jan 2001) by Mr James Testa, Director Editorial Development, they (ISI) are working on "forging full text links from the Web of Science to the journals appearing in the SCIELO Project database" (see [45] ). This "anchoring" of the locally created databases to the well-established international information centres has become a realistic prospect in ameliorating the South-North scientific information exchange.

The other possible ("second order") linkage, that of local citation indexes, is also in sight. The Polish Sociology Citation Index seems to have become quite viable within two years [38] [46] . While this example of social and humanistic coverage of scientific literature is probably the most needed "addition" to the existing ISI publications, others, offering to surmount the language barriers are appearing, according to Mr Testa's information: "In 1999 I …received a copy of the Chinese S&T Journal Citation Reports.  In the meantime, the Brazilians have also developed a Journal Citation Reports based on their own journals."

So, the information technology impact [1] gains in strength while this review was being compiled. Perhaps instead of helping to increase the worldwide visibility of selected journals from the science periphery the present review is already a piece for documenting the history of part of science communication. For some time to come, still, a common methodology for evaluating the scientific communicability of journals from science periphery might be useful in local science policy decision making.

Acknowledgements

This review could not be accomplished without the kind hospitality (in a study kernel with Internet facilities) of the National and University Library Zagreb, Croatia. It is a pleasure for the author to express his gratitude to the Director General Dr. Josip Stipanov.

Thanks are due to all those colleagues who helped by providing books and some reprints of papers otherwise unavailable (in time) to the author, as well as many a useful opinion: Mr. Juan Carlos Anduckia, Professor Ana Maria Cetto, Dr. Zvjezdana Dukia, Dr. Yuri Jack Gomez, Dr. Vesna Luzar-Stiffler, Dr. Penelope Murphy, Professor Linda Richter, Mrs. Iona Robu (M. Sci.), Professor Ronald Rousseau, Dr. Jane Russel. While the author is responsible for all the bad sides of this review thanks deserve those who cared to comment the first draft, and especially Dr. Gomez with respect to the notion of "evaluation indeterminacy".

References

[1] The Global Research Village: How Information and Communication Technologies Affect the Science System. Online http://www.oecd.org//dsti/sti/s_t/scs/prod/Global.pdf

[2] The International Encyclopaedia of the Social & Behavioral Sciences (Fall 2001) http://www.iesbs.com/sample.html (See "Scientometrics" by Loet Leydesdorff.)

[3] Manten AA: Publication of scientific information is not identical with communication Scientometrics 2/4 1980 303-308.

[4] Rabkin YM, Inhaber H: Science on the periphery: a citation study of three less developed countries Scientometrics 1/3 1978 261-274

[5] Stolte-Heiskanen V: Evaluation of scientific performance on the periphery Science and Public Policy 13/2 1986 83-88

[6] Maricic S: Evaluating Periodicals at the Scientific Periphery Iaslic Bulletin 38/1 1993 1-16

[7] Moravcsik MJ: Some contextual problems of science indicators. - In: Handbook of Quantitative Studies of Science and Technology AFJ van Raan (Ed.) Northholland, Amsterdam-New York-Oxford-Tokyo 1988 pp. 11-30

[8] Milic V: Recent trends in science and technology in developing countries Sociologija-Sociology 32/spec.issue 1990 351-626

[9] Strengthening the coverage of Third World science -The bibliographic indicators of the Third World´s contribution to science (Deliberations, Conclusions, and Initiatives of an ad hoc International Task Force for Assessing the Scientific Output of the Third World, 1985) The Final Report of the Philadelphia Workshop and of the Discussions Proceeding and Following that Workshop. Issued by Michael J. Moravcsik, Institute of Theoretical Science, University of Oregon, Eugene, Oregon 97403 USA February 1986. (16 pp.)

[10] Gibbs WW: Lost science in the Third World Scientific American August 1995 76-83

[11] Canagarajah AS: 'Nondiscursive' requirements in academic publishing, material resources of periphery scholars, and the politics of knowledge production Written Communication 13/4 1996 435-477

[12] Gaillard J: Science Policies and Cooperation in Africa Knowledge: Creation, Diffusion, Utillization 14/2 1992 212-233 (note 9)

[13] Russell JM: Publication Indicators in Latin America revisited. In: The Web of Knowledge: A Festschrift in Honor of Eugene Garfield. Edited by: Blaise Cronin and Helen Barsky Atkins . American Society for Information Science/Information Today, Medford, New Jersey, 2000, pages 233-250.

[14] Penelope Murphy: Determining Measures of the Quality and Impact of Journals (Commissioned Report No. 49, August 1996, National Board of Employment, Education and Training/Abridged Report Prepared for the Australian Research Council) Australian Government Publishing Service, Canberra, 1996.

[15 ] Pouris T & Richter L: Investigation into state-funded research journals in South Africa South African Journal of Science 96/3 2000 98-104

[16] Narin F: Objectivity vs. relevance in studies of scientific advance Scientometrics 1/1 (1978) 35-41

[17] Maricic S: A fourfold bibliometric evaluation of journals from less developed scientific communities - the pattern of references in medical journals from Yugoslavia Annals of Library Science and Documentation 30/2 1983 45-62

[18] Maricic S: Evaluating the scientific merit of journals within the scientific periphery; pp 61-74 in Kuenhold WW and Kirchner WP (eds.): Publications as an integral part of scientific research. Proceedings of the 5 th international conference of scientific editors. Hamburg, June 14-19, 1987, FR Germany. Published by the Bundesforschungsanstalt fuer Fischerei, Information/Dokumentation, on behalf of the Int. Fed. Scient. Editors' Ass., Hamburg 1988

[19] Maricic S: Evaluating Yugoslav (biomedical) journals Periodicum biologorum 92/2 1990 229-236

[20] Moravcsik MJ: Applied scientometrics: An assessment methodology for developing countries Scientometrics 7/3-6 1985 165-176 (This is an extended summary of a background report of some 70 pages prepared by MJM for a workshop held in Graz, Austria, May 2-7 1984, as indicated in ref. 5 of this paper.)

[21] Moravcsik MJ: Assessing the methodology for finding a methodology for assessment Social Studies of Science 16/ 1986 534-539

[22] Cetto AM & Hillerud KI (compiladores): Publicaciones Científicas en América Latina / Scientific Publications in Latin America México DF: ICSU.UNESCO.UNAM.AIC.FCE. 1995

[23] Cetto AM & Alonso O (compiladores): Revista Científicas en América Latina / Scientific Journals in Latin America (pp. 245-257). México DF: ICSU.UNAM.CONACYT.FCE. 1999

[24] Dantan AG: XVIII. ?Es posible evaluar cualitativamente las revistas academicas? Pp.187-199 in [22]

[25] Dantan AG: XXVII. El editor de publicaciones periodicas academicas, ?Una especie en peligro de extinction? Pp. 365-374 in (Cetto and Alonso, 1999).

[26] Vessuri H: Recent strategies for adding value to scientific journals in Latin America Scientometrics 34/1 1995 139-161

[27] Cetto AM: Scientific journal publishing in the developing world? COSTED (Committee on Science and Technology in Developing Countries, International Council for Science) Occasional Paper No.3 July 1998

[28] Gaillard J: Use of publication lists to study scientific production and strategies of scientists in developing countries Scientometrics 23/1 1992 57-73

[29] Guédon J-C: Beyond Core Journals and Licenses: The Paths to Reform Scientific Publishing ARL Bimonthly Report 218 October 2001 pp. 1-8 accessible at http://www.arl.org/newsltr/218/guedon.html (A full argumentation is at http://www.arl.org/arl/proceedings/138/guedon.html )

[30] Maricic S: Mainstream-periphery science communication Learned Publishing 13/4 2000 266 (Accessible at http://www.alpsp.org.uk/volcont.htm ( [18] ends with the sentence in which this idea was propounded for the first time, 15 years ago.)

[31] Maricic S: The missing link - The mainstream-peripheral science communication Current Science [India] 75/5 1998 427–428

[32] Gomez YJ, Anduckia JC, Rincon N: Publicaciones seriadas cientificas colombianas Interciencia 23 /4 1998 208-217

[33] Robu I, Marineanu D, Aciu I, Wood-Lamont S: Improving standards in the scientific biomedical community in Romania by using journal ranking to improve journal quality Health Information & Libraries Journal 18/2 2001 (in print)

[34] Maricic S, Sorokin B, Papes Z: Croatian journals at the end of the 20 century. A bibliometric evaluation Drustvena istrazivanja 9 /1 2000 1-18

[35] Gomez YJ: XXVIII. A proposito de un ejercicio de evaluacion de publicaciones seriadas cientificas. Pp. 375-390 in [23]

[36] Morris S: Learned journals and the communication research Learned Publishing 11/4 1998 253-258

[37] Smith LD, Best LA, Stubbs AD, Johnston J, Archibald AB: Scientific graphs and the hierarchy of the sciences - A Latourian survey of inscription practices Social Studies of Science 30/1 2000 73-94

[38] Winclawska BM: Polish sociology citation index (Principles for creation and the first results) Scientometrics 35/3 1996 387-391

[39] Stegmann J: Building a list of journals with constructed impact factors Journal of Documentation 55/3 1999 310-324

[40] Turban E, Zhou D, Ma J: A Methodology for Evaluating Grades of Journals: A Fuzzy Set-based Group Decision Support System Proceedings of the 33rd Hawaii International Conference on System Sciences - 2000 0-7695-0493-0/00 2000 IEEE -

[41] Moravcsik MJ: The coverage of science in the Third World - The "Philadelphia Program" in INFORMETRICS 87/88 edited by Leo Egghe and Ronald Rousseau, 147-155, Elsevier, Amsterdam, 1988.

[42] Cetto AM, Alonso-Gamboa O: 7 Scientific and scholarly journals in Latin America and the Caribbean, in Knowledge dissemination in Africa: The role of scholarly journals eds. Altbach Ph-G, Teferra D, Bellagio Publishing Network, Boston Massachusetts 1998

[43] Cetto AM, Alonso-Gamboa O: Scientific periodicals in Latin America and the Caribbean - a global perspective Interciencia 23/2 1998 84-93

[44] http://www.dgbiblio.unam.mx and http://www.latindex.unam.mx

[45] http://www.scielo.br

[46] Webster B: Polish Sociology Citation Index as an example of usage of national citation indexes in scientometrics Journal of Information Science 24/1 1998 19-32