Cumulative literature

Cumulative as an adjective is something that is increasing in amount, in effect or in strength with each successive addition.  

Hedges (1987, p. 443) writes that cumulativeness in the scientific enterprise can mean at least two things. In the broadest sense scientific results are cumulative if empirical laws and theoretical structures build on one another so that later developments extend and unify earlier work. This idea might be called conceptual or theoretical cumulativeness. The assessment of theoretical cumulativeness must be rather subjective. A narrower and less subjective indicator of cumulativeness is the degree of agreement among replicated experiments or the degree to which related experimental results fit into a simple pattern that makes conceptual sense. This might be called empirical cumulativeness.
 

Hedge further writes (1987, 453) "From a constructionist point of view, consistency of research results implies either the stability of the social constructions across the contexts in which experiments were conducted or an interpretive norm that leads to the perception of consistency. Thus the most interesting object of study surrounding research results that are perceived to be consistent may be the interpretative agreements that make possible the perception of consistency . . . One of the most interesting implications of the constructionist perspective is that the perceived cumulativeness of any research domain is a function of the conventionens of evidence and methodology in the research community".  

I general is the scientific literature considered cumulative, while the humanist literature is not. There exists, however, examples of the opposite way of expressing this relation. Langridge (1976, p. 30) thus writes:

"The respective products of the sciences and humanities are also markedly different. Science is sequential, producing what are regarded as successively closer approximations to scientific truth. The literature of science therefore has a limited life-span, and sooner or later is relegated to the history of science from the current body of scientific knowledge. The humanities by contrast are cumulative; Plato exists today alongside Kant, Whitehead and A. J. Ayer; in so far as each age has to recreate the past for itself....." Langridge (1976, p. 30)

Refinetti (1989) writes:

"For centuries, philosophers have discussed whether knowledge progresses analytically or dialectically. In the Cartesian tradition of starting with simple concepts and then building up more complex concepts [4], the idea of science as a gradual accumulation of small pieces of knowledge was put forward by Auguste Comte [5]. This constitutes an analytical view of progress of knowledge. On the other hand, Hegel proposed that knowledge grows as a whole, so that contradictions between opposing ideas are solved (and disperse pieces of knowledge are integrated) at each stage of the dialectical progress of knowledge [6]. This constitutes a dialectical view of the progress of knowledge. The partial correctness of both the analytical view of Comte and the dialectical view of Hegel have been acknowledged for many years...". (Refinetti, 1989).

Literature:

Gregersen, F. & Køppe, S. (1985). Videnskab og lidenskab. København: Tiderne skifter.
 

Hedges, L. V. (1987). How Hard is Hard Science, How Soft is Soft Science? The Empirical Cumulativeness of Research. American Psychologist, 42, 443-455.

Hägerstrand, T. (1985), The identification of progress in learning, Cambridge: Cambridge University Press. This series of essays discusses how progress is identified in a range of disciplines - physics, mathematics, biology, medicine, sociology, linguistics, art history, history, economics and ecology. The articles are based on discussions at a symposium organised by the European Science Foundation and are by acknowledged leaders in the fields covered. They seek to promote communication between the different disciplines, to identify the criteria of advancement and to examine problems in assessing them.
 

Kochen, M. (1974). Integrative Mechanisms in Literature Growth. London: Greenwood Press.(Contributions in Librarianship and Information Science Number 9).
 

Langridge, D. W. (1976). Classification and Indexing in the Humanities. London: Butterworths.
 

Rapoport, A. (1986). Examining the Concept of Advances, Especially in Psychology. (Chapter 14 in: Advances in the Social Sciences, 1900-1980. What, Who, Where, How? Edited by Karl W.Deutsch, Andrei S. Markovits & John Platt. Lanham: University Press of America, p. 287-307).
 

Refinetti, R. (1989). Information Processing as a central issue in philosophy of science. Information Processing & Management, 25, 583-584.
 

Scotchmer, S. (2005). Standing on the shoulders of giants: Protecting cumulative innovators. IN: Innovation and Incentives,  MIT Press.   Available: http://ist-socrates.berkeley.edu/~scotch/ch5.pdf

See also: Hard science versus soft science; Progress (Epistemological lifeboat).

Birger Hjørland

Last edited: 05-05-2006

Home

to be edited:

1) At en bibliografi - f.eks. Dansk Bogfortegnelse - er kumulativ vil sige, at nye og ældre informationer integreres i een struktur. Hvis en bibliografi f.eks. har systematisk orden og bl.a. alfabetiske registre, så ordner de kumulerede udgaver de samme referencer ordnet efter samme systematik og med samme registre. Da det forudsættes, at man ikke har behov for også at have disse bibliografiske informatio­ner organiseret efter optagelsestidspunkt i bibliografien, så overfløddiggør de kumulerede bind de hidtidige udgaver. Sådanne kumulerede udgaver kan fremstilles edb-teknisk. Vi vil derfor betegne denne form for kumulativ litteratur for mekanisk kumulering.

2) I henhold til den induktive og positivistiske videnskabsteori består forskningsprocessen i en indsamling af flere og flere observa­tioner, der giver anledning til mere og mere omfattende teorier. Det ligger således tæt op ad disse videnskabsopfattelser at sammenligne den videnskabelige proces med en mekanisk kompilation af bibliografiske data. Heri ligger også en opfattelse af den videnskabelige litteratur som havende kumulative egenskaber, der minder om den mekaniske kumulering. Megen informationsvidenskabelig forskning har bygget på sådanne - for det meste u­reflek­terede - positivistiske antagelser om den videnskabelige *informationsstruktur, jfr. f.eks. obsolesence.

3) I henhold til andre videnskabsteoretiske opfattelser, herunder Kuhn's paradig­meteori, har den videnskabelige proces ikke en sådan kompila­to­risk karakter, men er styret af teorier og hypoteser, der endog kan være ubevidste.


Den videnskabelige litteratur har således ikke kumulerende karakter i den mekaniske forstand, hvorimod faglig littertur kan have kumulerende karakter ("growth of knowledge") i meget forskellig betydning, form og omfang. En vigtig forudsætning for kumulativ vækst er, at forskning bygger videre på et bestemt paradigmes teorier og begreber. Man må altså vælge et paradigme og forholde sig tro overfor dets grundlag, og kun fravige paradigmets grundlag, hvis tvingende grunde gør dette nødvendigt. Disse tvingende grunde må da behandles meget eksplicit, og man må bevise, at afvigelsen fra paradigmet ikke blot skyldes forskerens eget ukendskab til det.

Et kriterium på, at en litteratur er kumulativ kan således være, at den optager tidligere litteraturs erkendelser i sig og dermed overfløddiggør denne. Når problemet flyttes fra den mekaniske til den intellektuelle kumulering, kommer problemstillingen til at indeholde forskellige erkendelsesteoretiske problemer. I det omfang enkelt-undersøgelser f.eks. bidrager til stadig mere almen, generel viden på et område, da vil der være tale om kumulative tendenser. Så filosofi­ske begreber som almengørelse og abstraktion er relevante for at forstå den videnskabelige litteraturs kumulative træk.

Informationsvidenskabens interesse i begrebet kumulativitet ligger bl.a. i dens forsøg på at formidle integrerende viden, ikke blot disintegrerende videns­fragmenter. Dette problem har f.eks. Kochen (1974) beskæftiget sig med.


Den videnskabelige litteraturs kumulering i videnskabsteoretisk lys behandles bl.a. hos Gregersen & Køppe (1985, 197-209).
 

the role of theories