College and Research Libraries WILLIAM E. McGRATH, DONALD J. SIMON, AND EVELYN BULLARD Ethnocentricity and Cross-Disciplinary Circulation Student circulation of books in forty-three major academic disciplines were examined for patterns of disciplinary interdependence. Percentage of books charged out by majors in their own discipline was defined as the ethnocen- tricity of the major. Percentage of books in a discipline charged out by stu- dents majoring in other disciplines was defined as the supportiveness of that discipline. The two concepts hatJe little or no correlation with each other. Graduate students were more ethnocentric than undergraduates. Most dis- ciplines were less supportive at the graduate level. Findings hatJe implica- tions for collection detJelopment . - uNIVERSITY STUDENTS use books in a wide variety of subjects, including those relating not only to their own major , but to other majors as well. For example, physics and chemistry students use books on mathemat- ics, and political science students use books on history and sociology, and so on. Not all such patterns are well known nor are they explicitly built into library policy and pro- cedure. In a university with sixty or more academic major departments, where it is conceivable that a student majoring in one could use books relating to every other, the number of such relationships becomes hor- rendous-so many, in fact, that even though librarians and faculty often say that these re- lationships should be considered in collec- William E . McGrath is dean of library ser- vices, University of Lowell , Lowell , Massachu- setts. Donald ]. Simon is director of information systems, and Evelyn Bullard is circulation librar- ian , University of Southwestern Louisiana , Lafayette . This article is a revision of a paper by William E. McGrath , " Disciplinary Interdepen- dence in University Students' Use of Books, " presented at the annual meeting of the American Society for Information Science, 1976, and first published in Information Politics (Washington , D.C.: American Society for Information Science , 1976), p.98, microfiche 6. tion building, the problem is generally ig- nored, and solutions are not sought. This paper seeks to uncover some of these patterns in one university by examin- ing (1) the extent to which students major- ing in one subject will use the books of another subject and (2) the extent to which the books in one subject are read by stu- dents majoring in other subjects. A compos- ite and economical picture of this discipli- nary interdependence as it pertains to stu- dent library use has never been painted. The term disciplinary interdependence is used rather than the term interdisciplinar- ity. The first term retains the identity of older established disciplines, such as physics and mathematics , while measuring their de- pendence on each other, whereas interdis- ciplinarity refers to newer disciplines, such as biochemistry , that have an identity in their own right. · This interdependence should be much broader at the undergraduate level when students have not yet fully specialized and · are required to read broadly. At the gradu- ate level, specialization should have nar- rowed the breadth of dependence while in- tensifying the depth. Knowledge of this interdependence should have implications for library opera- tions (circulation control, allocations, acqui- 511 512 I College & Research Libraries • November 1979 sitions, etc.), and in terms of political com- petition for library funds and priorities. It should also be a measurable reflection of course enrollments and assignments. THEORETICAL CONTEXT This research can be placed in the context of Merton 's definition of disciplines as (a) bodies of knowledge and (b) organizations of practitioners . 1 Here, the "bodies of knowl- edge" are all the subject fields embraced by each academic department in a specific uni- versity . "Organizations of practitioners" are groups of students majoring in the fields of those same departments. The research is also of interest in Kuhn- ian terms-e. g. , differences in textbook use , the research frontier , and paradigm development. 2 Although Kuhn speaks primarily of the scientist and how that per- son uses books and journals, the scientist's behavior should be reflected by students to the extent they use the same books and journals. Graduate students, i.e., those who are more specialized than undergraduates, presumably do more research than under- graduates. Their use of books , accordingly , should reflect this specialization. Sherif and Sherif comment that "each dis- cipline needs others in a fundamental and basic sense as a validity check on its own generalizations and theories ," and that "man does not arrange his problems along lines drawn by academic disciplines. "3 Donald T. Campbell expands on this ob- servation in his concept of enthnocen- tricity-the tendency of practitioners to ig- nore knowledge outside of their own disci- pline.4 The tendency , of course, should vary from discipline to discipline. Campbell ad- vocates the ideal situation , what he calls the " fish-scale" model of omniscience-i.e. , each discipline overlaps those adjacent to it. He contrasts this fish-scale view with the present situation in which disciplines over- lap within clusters , in isolation from other clusters. He proposes the fish-scale ap- proach to graduate training . The study undertaken here offers one approach to measurement of this concept. As a measure of Merton ' s bodies of knowledge , the term supportiveness is de- fined here as the extent to which books on the subjects taught in a university depart- ment are used by students majoring in other subjects. It might be supposed that the term basic could be appropriately used here . However, basic usually refers to re- search by scientists, whereas supportiveness is defined in the context of the university curriculum. PURPOSE The purpose of this research is to con- tribute to the understanding of library use in terms of all majors on the one hand , and all subjects on the other-that is , who uses what and what is used by whom. The two phrases are not the same , as will be seen. Use by an individual person is not exam- ined. Use is aggregated by subject arid major, and the identity of individuals is lost, so that we do not know which specific books were used by any specific individual. In other words, the study is sociological rather than psychological in that the group-i.e ., the major itself, rather than the person-is the unit of interest. "All subjects" in this paper refers to the subject matter described by the courses listed under each academic department in the University of Southwest- ern Louisiana ( U.S.L. ) Bulletin. " All majors'' refers to departments granting at least a bachelor's degree. The University of Southwestern Louisiana is accredited by the Southern Association of Schools and Colleges. It awards degrees at the bachelor , master's , and · Ph. D. levels , with an enrollment of approximately 11,000 students and a library of nearly 500,000 vol- umes. It is located in the heart of French Louisiana, which contributes much to the color and tradition of the university . It has advanced research programs in biology, his- tory , English literature, microbiology, mathematics , and computer science. As usual in circulation studies , use is as- sumed when a book circulates. For what purpose a book may be used is not consid- ered here. Specific questions considered were: 1. To what extent do graduate and undergraduate students use books on sub- jects confined to their own major? That is , to what extent is the major ethnocentric? 2. Which subject areas are most support- ive in that they are most heavily used by graduate and undergraduate students in other disciplines, and to what extent? That is, to what extent does the subject matter of a discipline support majors in other disci- plines? 3. What differences exist between gradu- ate and undergraduate use? This question can be expressed by four null hypotheses: that (a) no correlation exists between gradu- ate and undergraduate ethnocentricity; (b) no correlation exists between graduate and undergraduate supportiveness; (c) no dif- ference in mean percentages exists between graduate and undergraduate ethnocentricity; and (d) no difference in mean percentages exists between graduate and undergraduate supportiveness. We would expect hypoth- eses (a) and (b) to be retained and (c) and (d) to be rejected, supporting the contention ~hat differences exist between the two levels. Furthermore, we would expect the mean percentage for ethnocentricity to be higher at the graduate level and the mean percentage for supportiveness to be lower at the graduate level. 4. What is the difference between ethnocentrity and supportiveness? This question can be expressed by two null hypotheses: that no correlation between the two exists at (a) the graduate level and (b) the undergraduate level. We would expect these hypotheses to be retained. PREVIOUS RESEARCH Previous research along these lines, sometimes called "dispersion of the litera- ture," has usually been confined to citation studies of scientific journals . Earle and Vickery, for example, counted citations in journals from various disciplines to deter- mine the extent each cited or was cited by other disciplines. 5 Moore classified scientific journals accord- ing to the Dewey classification system, then ranked them according to how much the journals were confined to each subject cate- gory and how much they were assigned sub- jects related to other categories. 6 Narin, Carpenter, and Berlt studied cross-citing among 275 journals in mathe- matics, physics, chemistry, biochemistry, and biology and concluded from citation patterns that these sciences were transi- tively related to each other in that order. 7 Such studies are abundant. These authors Ethnocentricity I 513 can recall no similar empirical studies that have examined the cross-disciplinary use of monographs by students . Baughman, how- ever, has argued for the study of interrela- tionships within and among clusters of sub- ject literatures, demands, and disciplines in building library collections. 8 METHODS Student borrowers were classified into forty-three academic major areas. These areas are official academic departments of U.S. L. They are also among the disciplines recognized in the U.S. Office of Education, Higher Education General Information Sur- vey (HEGIS). 9 Monographs circulated to the students were grouped into the same forty-three academic areas, according to the books' clas- sification numbers, using the method de- veloped by McGrath and Durand. 10 Two computer programs were written to process yearly circulation on U.S.L.'s UNIVAC SPECTRA 70/45. The first, written in COBOL, processes a tape of book charges compiled from the li- brary's IBM 357 data input system, using a file of student I. D. numbers, majors, and class years. It groups each student into one of the HEGIS categories according to his or her major, and each book's classification number into its proper HEGIS category. For each charge, the student's major and the book's subject area are not necessarily the same. The program separates charges according to whether they are graduate or undergraduate . The second program in PU1 arrays and prints the number of book charges into two rectangular, nonsymmetric matrices, one for graduates and one for undergraduates. The columns consist of academic majors (per- sons) and the rows of academic subjects (bodies of knowledge). Each cell of the ma- trix contains the number of books charged according to the subject of the row and the major of the column, so that the number of books charged in any . subject by any major group is determined. Four matrices for two years' circulation, 1974/75 and 1975/76, were constructed, two for undergraduates and two for graduates. Two years' data were collected since circula- tion patterns may change over time. The 514 I College & Research Libraries • November 1979 matrices are too large to include in this paper. Instead, an excerpt with typical data is shown in table 1. One test for the validity of subject classi- fication can be made by examining the numbers in the diagonal. Students should ordinarily borrow more books in the subject of their own major than in other subjects. If the data showed otherwise, subject classi- fication of departments should be suspect. Disciplines selected as examples for tbe matrix in table 1 all have large diagonal cells. For undergraduate diagonal cells, nine- teen of forty-three in 197 4/75 were largest and twenty-eight of forty-three in 1975/76. In nearly all cases where the diagonal cells were not the largest, English literature was larger, indicating heavy dependence by all majors. For graduate diagonal cells, sev- enteen of eighteen in 1974/75 were largest, and sixteen of nineteen in 1975/76. These ratios would suggest substantial validity of classification. Unit of analysis in this study was the academic major or discipline. Variables were ethnocentricity and cross-disciplinary support (supportiveness). Scores for ethnocentricity were percentages obtained by dividing each diagonal value by the total in its respective column. Scores for support- iveness were percentages obtained by divid- ing each row total, less its diagonal value, by the row total. Percentages were obtained for all disciplines for both academic years, 1974/75 and 1975/76. Pearson product-moment correlations were then used to test the similarity of the percentages for the two years. Two-year correlations for undergraduate ethnocentric- ity, graduate ethnocentricity, undergraduate supportiveness, and graduate supportiveness were 0.83, 0.85, 0.91, and 0.80 respec- tively. These correlations were high enough to indicate little change of circulation pat- terns over two years~ All subsequent analysis, therefore, was done on the aver- ages for the two years. Pearson correlations were also used to test the relationship between the two con- cepts of ethnocentricity and supportiveness. Spearman rank correlation was used to test shifting of ranks between undergraduate and graduate majors. Student's t-test was used to test the difference between means of undergraduates and graduates. RESULTS Ethnocentricity Table 2 shows results relating to question 1, the extent that undergraduate students use books on subjects in their own major. Music students, for example, borrowed 924 books about music in 1974/75, but they bor- rowed a total of 1,278 books on all subjects, or 72.3 percent. Similarly, in 1975/76, they borrowed 71.1 percent in their own field. The average of these two percentages was 71.7. Majors are ranked according to this aver- age percentage. The higher the percentage, the more ethnocentric the major. That is, students in high ethnocentric disciplines read in their own subject more than they do in other subjects. The lower the percentage, i.e., the less ethnocentric, the more reading they do in subjects other than their own major. Among undergraduates, music TABLE 1 Subjects Biology Chemistry Computer Science Education French History Mathematics Music Psychology Speech CIRCULATION MATRIX OF SUBJECfS AND MAJORS , GRADUATE STUDENTS, 1975/76, EXCERPT SHOWING FORMAT Computer Majors Biology Chemistry Science Education French History Mathematics 450 3 32 68 1 3 3 15 102 3 22 0 1 4 0 0 731 3 0 0 2 0 0 28 1,672 4 5 5 0 0 0 96 653 12 2 28 5 11 256 45 862 1 35 3 186 45 0 2 561 2 8 10 264 1 5 0 18 0 2 534 10 11 1 0 0 0 82 0 2 0 Music Psychology Speech 0 54 2 0 2 0 0 0 9 0 11 41 0 0 0 0 17 6 0 32 5 166 11 4 2 483 90 0 1 375 TABLE 2 ETHNOCENTRICITY, UNDERGRADUATE LEVEL: PERCENT OF BOOKS ON ALL SUBJECTS CHARGED BY STUDENT MAJORS IN THEIR OWN MAJOR SUBJECT; RANK ORDER OF AVERAGE PERCENT BY MAJOR Rank Stude nt Majors Two-Year Average Pe rce nt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Music English History Electrical Engineering Civil Engineering Spanish Architecture Home Economics Microbiology Geology Petroleum Engineering Mathematics Psychology Fine Arts Horticulture Political Science Speech Sociology Philosophy Computer Science Applied Arts French Chemical Engineering Mechanical Engineering Economics Biology Physics Agriculture Nursing Industrial Arts Accounting Medical Records Marketing Special Education Education Finance Chemistry German Journalism Management Geography General Business Vocational Education • Most e thnocentric. t Least e thn oce ntric . 71. 7* 44.6 42.6 41.6 36.8 34.6 33 .8 29.6 28.7 25 .9 24.5 23.3 23.0 21.8 21.2 21.1 21.0 20.0 19.4 18.6 18.3 17.9 17.1 16.7 16.5 15.9 12.9 12.8 ll.8 9.4 8.7 8.4 8.3 6 .9 5.3 5 .1 5.1 4 .5 3.3 3.2 2 .8 1.3 o.ot majors are most ethnocentric; geography, general business, and vocational education majors are least ethnocentric. These results are, of course, for aggregate use . Patterns of individual use may well be different. Table 3 shows the extent that graduate students charged out books in their own major and their ranks . Here music js again Ethnocentricity I 515 ranked highest in ethnocentricity and man- agement lowest. Supportiveness Table 4 shows the data relating to ques- tion 2: the extent that undergraduate sub- jects are supportive of major areas other than their own. Taking music again as an example, 70.3 percent of books on music were charged out in the two years by undergraduate nonmusic majors. The subjects are ranked according to the two-year average percentages of books taken out by nonmajors. A higher percentage in- dicates less book use in a subject by stu- dents majoring in that subject, and more by students majoring in other subjects. That is, the higher the percentage, the more that subject supports other academic subjects. For example, German and vocational educa- tion ranked highest in supportiveness with nearly 100 percent of the books in these subjects checked out by other majors. Nurs- ing was least supportive, with 24.7 percent. Table 5 shows the extent that graduate subjects are supportive of other major areas and their ranks. Here, management was most supportive and computer science the least. Rank 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 TABLE 3 ETHNOCENTRICITY, GRADUATE LEVEL: PERCENT OF BOOKS ON ALL SUBJECTS CHARGED BY STUDENT MAJORS IN THEIR OWN MAJOR SUBJECT; RANK ORDER OF A VERAGE PERCENT BY MAJOR Stude nt Two-Year Ave rage Majors Pe rcent Music 87.4 Spanish 85.0 English 74.3 Mathematics 71.6 Sociology 65.3 French 64.1 Psychology 55.8 Chemistry 49.7 History 48.2 Computer Science 45.7 GeoloW 38.8 Speec · 36.5 Home Economics 32.8 Microbiology 32.2 Ph~sics 31.0 Po 'tical Science 25.0 Geography 23.7 Education 23 .3 Management 2.2 516 I C allege & Research Libraries • November 1979 Undergraduate and Graduate Differences Results from question 3, the difference between undergraduate and graduate use, are shown in tables 6 and 7. Considerable shifting of ranks from the undergraduate level to the graduate level for both ethnocentricity and supportiveness is appar- ent, as the correlation coefficients in table 6 show. For example, whereas undergraduate TABLE 4 SUPPORTIVENESS , UNDERGRADUATE LEVEL: PERCENT OF TOTAL BOOKS CHARGED IN EACH S UBJECT BY NONMAJORS IN THE S UBJECT; RANK ORDER OF A\ ERAGE PERC ENT Rank 1 2 3 4 5 6 7.5 7.5 9 10 11 12 13 14 15 16 17 18 19 20 21 . 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 Subject Vocational Education German Geography Finance Philosophy Chemistry General Business History English Spanish Economics Industrial Arts French Physics Math. & Statistics Management Horticulture Biology Psychology Sociology Home Economics Applied Arts Speech Civil Engineering Special Education Agriculture Fine Arts Microbiology Marketing Journalism Mechanical Engineering Music Political Science Geology Chemical Engineering Computer Science Medical Records Tech. Education Petroleum Engineering Electrical Engineering Accounting Architecture Nursing *M ost supportiv e. tLeast supportive . Two- Year Average Pe rce nt 100.0* 99.8 96.0 94 .1 93 .8 93.4 93 .0 93.0 92 .8 92.6 91.8 91.6 90.2 89.5 88.9 88.0 84.7 84 .2 84.1 83.4 81.9 81.6 79.5 78.7 78.5 76.9 76 . 1 75.2 74.1 73 .0 71.0 70.3 66.0 64.3 59.9 54.5 42.5 39.1 38 .6 32 .8 27.2 26.1 24.7t TABLE 5 SUPPORTIVE ESS , GRADUATE LEVEL: PERCENT OF TOTAL BOOKS CHARGED IN EACH SUBJECT BY NONMAJORS IN THE SUBJECT; RANK ORDER OF AVERAGE PERCENT Two- Year Ave rage Rank Subject Pe rcent 1 Management 2 Home Economics 3 Physics 4 S~anish 5 C emistry 6 Music 7 Psychology 8 Political Science 9 Geology 10 Biology 11 History 12 French 13 Geography 14 Microbiology 15 Speech 16 English 17 Mathematics 18 Education 19 Computer Science *Mos t suppo rti ve. t Least supporti ve. 98.5* 97.9 85.0 82.8 78 .7 76.1 72.6 70.9 55 .5 55.2 53.8 47 .0 46 . 1 35.6 34.3 32.6 32. 1 19.7 13.lt biology majors ranked twenty-sixth of forty- three (table 2) in ethnocentricity, graduate biology majors ranked fifth of nineteen (ta- ble 3) . The shift for ethnocentricity was sub- stantial (r=0.53) and considerably greater for supportiveness (r=0.25) . Table 7 shows significant differences in the mean percentage levels for both ethnocentricity and supportiveness. The re- sults show the mean percentage for ethnocentricity to be twice as high at the graduate level. The results also show that most disciplines are considerably less sup- portive at the graduate level. Hypotheses 3(a) and 3(b) were thus re- tained , and 3(c) and 3(d) were rejected , as expected , supporting the contention that differences exist at the undergraduate and graduate levels. TABLE 6 CORRELATIO BETWEEN UNDERGRAD UATE AND GRADUATE LEVELS , SPEARMAN RANK ORDER COEFFICIENTS Ethnocentricity Supportiveness Correlation Coeffi cient u mber 0.53 0.25 18 18 Ethnocentricity versus Supportiveness Results for question 4 are shown in table 8. The virtual zero coefficients ( -0. 03, undergraduate , and -0.05, graduate) indi- cate little and probably no relationship be- tween the two concepts. Thus , hypotheses 4(a) and 4(b ) were retained as expected. DISCUSSION Graduate student reading was more ethnocentric than undergraduate , in that graduate students showed a higher percent- age of reading in their own subjects. This supports the commonly held belief that graduate reading is more specialized. Inter- dependence is indeed broader at the under- graduate le vel. Campbell, of course, as- serted that graduate students were too spe- cialized . Measurement of this specialization, as undertaken here, may provide a means for observing this specialization over time . It would be interesting to know whether students will be less or more specialized ten years from now. It is clear from the low correlation and different percentages that the two con- cepts-ethnocentricity and supportive- ness-are not the same and in fact are very different. One is a characteristic of the per- sons specializing in a discipline . The other is a characteristic of a discipline's relevance to persons outside of the discipline. The findings for ethnocentrism may also have relevance in terms of Kuhn's concept of paradigm development of disciplines , sometimes referred to as hard or soft. For example , if students make more use of books in their own major , that major may be farther along the continuum of paradigm development. Inversely, the more reading they do outside of their discipline , the less well developed the paradigm. This interpretation is cautious, since the patterns shown are not in agreeme:pt with TABLE 7 COMPARI SON OF U NDERGRAD UATE AND GRAD UATE LEVELS , DIFFERENCE BETWEEN MEAN PERCE NTAGES Unde r- graduate Graduate Difference Number Ethnocentricity 23.8 Supportiveness 80.2 47.0 57.2 23.2* 23.0* *Differences significant at O.Ollevel for 34df. 19 19 Ethnocentricity I 517 TABLE 8 CORRELATIONS B ETWEEN ETH NOCE NTRI C ITY AND S UPPORTI VENESS , PEARSO N PROD UCT MOMENT COEF F ICIENTS Undergraduate Graduate Corre lati on Coe ffi cie nt -0.03 -0.05 Num be r 42 18 studies showing paradigm development . For example, English , showing high ethnocen- trism in this study, is not regarded as a high paradigm discipline , in Kuhnian terms . Likewise , geology , showing low ethnocen- trism , would not be regarded as a low paradigm discipline. Relationship between ethnocentricity and paradigm development must remain a hypothesis , therefore. How much the findings for supportive- ness reflect theory and how much purely local conditions is very uncertain at this stage. A discipline showing high support- iveness for other disciplines may also be in- terpreted as having high self-supportive- ness. Findings to a large extent must be a function of the local curriculum , distribution of enrollment, and size of collection. In terms of aggregate use , the findings' have obvious implication and interest to campus politics (for both ethnocentrism and supportiveness) . In terms of competition for library funds and for building collections relevant to particular subject fields, faculty often argue that their teaching and research range beyond the immediate confines of their own discipline , often implying that this is not true of other areas and that allo- cations should take this into account. Pre- sumably , this argument holds for student use as well. As these findings show, it is in- deed true, and to a measurable degree, that library use , both by major and subjects of books used , ranges beyond the indicated discipline. The measured degree is large for some and small for others and forms a con- tinuum with disciplines all along the scale. The findings suggest that a better case can perhaps be made for allocation on the basis of total subject usage rather than on total major usage. After all, it is use made of the books in the collection that is of primary interest to collection builders. A third pos- sibility would be to allocate on the basis of use in a major's own subject (the diagonal in 518 I College & Research Libraries • November 1979 table 1), but this would tend to slight stu- dents who are interested in subjects other than their own and to arbitrarily restrict the range of book selection. Patterns of the kind discussed here, though the findings are merely indicative or tentative and not conclusive, may also be used to help determine other kinds of priorities. On the other hand, it can be ar- gued that existing university priorities- e.g., enrollment quotas, departmental em- phases, curriculum completeness-strongly influence what circulates from the library. Finally, no value judgments concerning ethnocentricity or supportiveness of stu- dents and faculty are intended here, nor should any be inferred from either the question or the results. Numerical values for these concepts, of course, are specific to the university studied because of its enroll- ment distribution and cannot be readily generalized from the data given here. "Ac- ceptable" levels of ethnocentricity and supportiveness-high, low, or otherwise- do not exist and probably should not. On the other hand, we can hypothesize that the relationship, or rather the dif- ferences, between undergraduate and grad- uate interdependence and the difference be- tween ethnocentricity and supportiveness also hold at other institutions. FURTHER RESEARCH Findings for supportiveness might be more generalizable if the "main effects" of the matrix-i.e., the row and column means-were subtracted from each cell , thus offering a means of comparing one en- vironment to another. This correction m.,_y also throw light on the paradigm question. Correlation studies may be done on the re- lationship between paradigm development and ethnocentricity. A multidimensional scaling of the entire matrix of majors and subjects is being undertaken. This analysis should identify clusters of subjects and clus- ters of majors and should measure both the degree of dependence of each discipline within a cluster and the distance between " clusters. This would provide another test of Campbell's model. REFERENCES 1. Robert K. Merton, The Sociology of Science: Theoretical and Empirical Investigations, edited with an introduction by Norman W. Storer (Chicago: Univ. of Chicago Pr., 1973). 2. Thomas Kuhn , Structure of Scientific Revo- lutions (Chicago: Univ. of Chicago Pr., 1962). 3. Muzafer Sherif and Carolyn W. Sherif, "In- terdisciplinary Coordination as a Validity Check," in their Interdisciplinary Relation- ships in the Social Sciences (Chicago: Aldine, 1969), p.3-20. 4. Donald T. Campbell, "Ethnocentrism of Dis- ciplines and the Fish-Scale Model of Omni- science," in Muzafer Sherif and Carolyn W. Sherif, eds., Interdisciplinary Relationships in the Social Sciences, p.328-48. 5. Penelope Earle and Brian C. Vickery, "Sub- ject Relations in Science!fechnology Litera- ture," Aslib Proceedings 21:237-43 (June 1969). 6. J. R. Moore, "On the Interrelationships of the Sciences and Technology as Expressed by a Categorized List of Journals and Mod- ified by a Classification System, " Journal of the American Society for Information Science 24:359-67 (Sept. 1973). 7. Francis Narin, M. Carpenter, and N. C . Berlt, "Inter-Relationships of Scientific Jour- nals," Journal of the American Society for In- formation Science 23:323-31 (Sept. 1972). 8. James C . Baughman, "Toward a Structural Approach to Collection Development," Col- lege & Research Libraries 38:241-48 (May 1977). 9. Robert H. Huff and Marjorie 0. Chandler, A Taxonomy of Instructional Programs in Higher Education (Washington, D.C.: U.S. Department of Health, Education and Wel- fare, National Center for Education Statistics, 1970). 10. William E . McGrath and Norma Durand, "Classifying Courses in the University Catalog," College & Research Libraries 30:533-39 (Nov. 1969).