Integrative Levels Classification project scheme monograph references

« Categories

Development of knowledge in time

Knowledge obviously changes in time. People in different epochs have different knowledge.

Supporters of relativistic views can even claim that knowledge of each epoch and culture is equivalent to that of any other, while no one is superior. The concepts developed by different cultures may additionally be considered as "incommensurable", that is, as belonging to completely independent systems that cannot be connected to each other.

What is true is that much knowledge of past cultures, like ancient Egyptians, or Incas, or Tibetans, is lost in such accidents of history as natural disasters, fires, or cultural repression by conquerors; only few fragments remain to us of the wisdom of pre-socratic Greek philosophers. Aspects of knowledge that had already been investigated in depth by some people, in their own languages and forms, have to wait for being considered and developed again several centuries later, in new languages and forms. Thus it is true that knowledge does not progress along a linear path, as it could be if it were all carefully kept, spread and studied in a cumulative way. This is an important reason why the work of knowledge organization is so much worth, as it helps to reduce knowledge wastes.

Still, the parts of knowledge that are conserved and trasmitted do develop in cumulative ways, by a "Lamarckian evolution" as some describe it. Progress exists, and has important practical consequences. Even skeptical thinkers would probably realize it more distinctly, if they had to eat only chestnuts for all the winter, or to be healed without penicillin...

Knowledge development means that any knowledge organization system cannot but reflect the state of knowledge at the time of its creation, or of its update, while no KOS can ever be definitive and absolutely valid throughout all times. Also the knowledge that we are representing in ILC is a provisional one, though hopefully more updated and structured than it was in pre-existing systems. In other terms, systems should be considered also in a diachronic dimension [Ten02], including the historical relations between different ways of conceptual organization. Such work will expectedly teach interesting and fruitful lessons on the development of knowledge in general, if one is willing to go beyond relativism.

Knowledge at a given epoch consists of concepts, that is, intellectual representations of real entities, as opposed to the real entities themselves [Dah74; Hjø09]. The cumulative system of concepts is an intellectual product of humans, laying at level y within the cultural stratum, or in World 3 in the terms of Popper, who emphasizes that knowledge lays at a level above both the material objects (World 1) and the individual minds processing sensory data (World 2) [*Pop72]. Thus, what we organize is not reality in itself (α), but concepts of it. Reality is not necessarily something mysterious and metaphysical, like Kant's noumenon, as at least some portion of it is reflected in phenomena; simply it is an external dimension prior to knowledge. One aspect of the difference between reality and concepts is that, while reality can be supposed synthetically as a whole, its knowledge is necessarily analytical, being articulated in single concepts and relations between them.

Conceptual knowledge involves several further dimensions: phenomena (β), aspects like theories and disciplines (γ), carriers (δ) etc. Thus, as the most basic units of knowledge we can take phenomena, that is reality as it is currently known by mankind (phenomenon etymologically means "manifest"). Phenomena can be said to be the elements of reality as far as we have been able to identify them. As such, although we aim to establish a durable list of them, they are subject to change as our knowledge develops.

Let us take a simple example from recent biological knowledge. Caracals are a class of felids that were denoted for longtime by the scientific name Felis caracal, as believed to be strictly related to the cats of the genus Felis, in the following phylogenetic chain:

mq	animals
mqv		chordates
mqvt			mammals
mqvto				carnivora
mqvtof					felids
mqvtoff						Felis
mqvtoffl						Felis caracal

Consequently, ILC would have represented them by the notation on the left. However, it has been found recently that the phylogenetic relationships of caracals are slightly different, so that they deserve to belong to a separate genus Caracal:

mq	animals
mqv		chordates
mqvt			mammals
mqvto				carnivora
mqvtof					felids
mqvtofr						Caracal
mqvtofrc						Caracal caracal

This also has the effect of sligthly modifying the ILC notation expressing their position: "caracals" are now mqvtofrc rather than mqvtoffl. Although caracals are a phenomenon, rather than a discipline or a theory practised at a given epoch, their classification is also subject to change. Clearly more abstract concepts, like theories and disciplines, belonging to the further dimension γ, are even more subject to change: while the phenomenon of caracals has only been moved slightly in its classification tree in several centuries, theories on caracals, such as morphology, physiology, or evolutionism have changed much more. This supports the idea of taking phenomena as a good basis for classification, despite they are not completely exempt from change. Actually, if our concepts would never change, this would mean that we know everything already from the beginning, hence do not need knowledge organization at all! At the opposite extreme, if we were completely unable to ever acquire any valid knowledge of phenomena, knowledge organization would be useless. Knowledge organization is important just because our knowledge or phenomena is valid in many parts but never perfect.

It can be observed that phenomena also contain some theoretical elements, like aspects do; or even that the existence of any phenomenon, eg the existence of caracals, is also a theory. However, knowledge tends to separate progressively these two components, that is, to consider as phenomena those concepts that have proved more stable and have not been falsified in a long time, while explicitly considering as theories those intellectual constructions that are under discussion against competing alternative constructions: while the existence of caracals, as a species of animals well distinct from other species, now seems to be an acquired fact, the model of punctuated equilibria in the evolution of caracals is still a theory. It can be noticed that some entities treated by advanced fields of knowledge, like current high energy physics, are more theoretical than others: Higgs boson is a theoretically predicted particle that can hardly be called a phenomenon, until more evidence of its actual existence is found [Szo04].

That knowledge develops in time, makes a single KOS facing two problems: how to represent outdated knowledge belonging to obsolete systems, and how to provide for future updates and changes of the current system. Although these problems challenge any KOS, some KOS are better than others in managing solutions to them. For example, hospitable notations make it easier to update the schedules while keeping a good representation of the order of concepts. This is made possible especially by decimality, as introduced by Dewey [Gre09], and by empty digits, as introduced by Ranganathan: both these devices are also adopted in ILC. Another common observation is that faceted KOSs are update-friendly, as new foci or new facets can be added where required, without changing the whole structure of the system. This does not guarantee that any change will always be possible. In case future knowledge realized that the model of integrative levels were a completely inadequate one, also the main structure of ILC would become unsuitable. What we can hope is that, by building a structure as more general and faithful to the actual relations between phenomena as possible, the system will require less substantial modifications in the medium term, and its usefulness will last for a longer time.

The first question posed by knowledge development is how to represent concepts that belong to outdated knowledge systems, such as "aether", "phlogiston" or "unicorns". As ILC is based on the current state of knowledge, in which these concepts have no place, their presence looks someway paradoxical. Still one may need to classify documents dealing with them, either as they follow an old knowledge perspective (eg a book of alchemy published in 18th century), or as they study these concepts in a historical perspective (an essay on the history of chemistry).

The meaning of an outdated concepts usually has a part which is outdated, and another which still makes sense in contemporary knowledge: aether, though not existing as such, was conceived as a material substance in the same way as existing material substances; unicorns, although fantastic, are conceived as mammals like existing cows or horses. Their representation in ILC will thus be possible by combining a standard part, eg g "bulk matter", with an aspect facet expressing the epoch of its knowledge, eg 01h "as known in 18th century"; further subdivisions may be defined to express the needed specific concepts: g01h)e "aether".

In case we want to adopt the same logic as that of the old time, we can define special classes, their subdivisions and facets in such a way that the original logic is reflected [*Gno09a]:

	g	substances
	gA 		earth
	gE		aether
	gF		fire

The second challenge is providing ways to manage the development of ILC itself, with its future updates. All major bibliographic classifications are published in successive editions incorporating each time a set of changes and additions. The number of the edition has even become part of the common way to refer to the systems, like in DDC22 for the the twenty-second edition of the Dewey Decimal Classification, or BC2 for the second edition of the Bliss Bibliographic Classification. Nothing prevents that, once a first edition of ILC will be stabilized, new editions or versions of it can be developed and replace the previous one. Publication on the Internet make it easier and quicker to distribute such updates.

Adopting a new edition is always problematic for users of the system that already have other resources classified with the old edition. Re-classifying is probably the most demanding task for librarians, especially in terms of time [Cro01] — which often means that the new edition is adopted only for new documents, leaving the old ones classified with the old edition. Also in this case, however, managing both the system and the resource metadata in a digital form gives big advantages. Equivalences (not only biunivocal, but sometimes more complex) between old and new notations can be recorded in a field of the database. In this way, it will then be easy to write algorithms to automatically shift from one edition to the other while searching or navigating, or to automatically update many classmarks in order to make them aligned with the new edition.

Arrays of levels »

References cited in this section

Cro01: Il rinnovamento della Dewey / Luigi Crocetti = Quali spazi per le classificazioni? : tavola rotonda AIB Piemonte : Torino : 20 gennaio 2001 ]] resoconto = (ISKO Italia. Documenti) — <> : 2004-

Dah74: Zur Theorie des Begriffs / Ingetraut Dahlberg = International classification. 1: 1974. 1. P 12-19

Gre09: Melvil Dewey's ingenious notational system / Rebecca Green = Proceedings North American Symposium on Knowledge Organization. 2 : Syracuse (NY) : 2009. P 91-99 / Elin K Jacob, Barbara Kwasnik = DLIST. 2635 — University of Arizona <>

Hjø09: Concept theory / Birger Hjørland = Journal of the American society for information science and technology. 60: 2009. 8. P 1519-1536

Szo04: Classifying science : phenomena, data, theory, method, practice. P 30 / Rick Szostak — Springer : Dordrecht : 2004

Ten02: Subject ontogeny: subject access through time and the dimensionality of classification / Joseph T Tennis = Challenges in knowledge representation and organization for the 21st century : proceedings of the Seventh international ISKO conference : Granada : 2002. P 54-59 / María José López-Huertas : ed' — Ergon : Würzburg : 2002


Integrative Levels Classification. Philosophy. Development of knowledge in time / Claudio Gnoli – ISKO Italy : <> : 2009.08.03 - 2011.07.29 -

  Integrative Levels Classification project scheme monograph references