There are boundary restrictions of who, what, and where on the information that can be made freely available to researchers, and these restrictions necessarily prevent the open Internet from providing “all” information freely to anyone, anywhere, at any time. These boundaries are created primarily by copyright laws and legal site license and password limitations that are just a much an unavoidable part of the information world as is computer technology itself. It is also the case that the higher levels of learning in our culture—knowledge, understanding, and wisdom as contrasted to data, information, and opinion—are conveyed primarily by book-length texts (either print or electronic in format) of sufficient amplitude to convey large perspectives integrating wide ranges of information into coherent frameworks of exposition or narration.
The means of discovering, evaluating, and bringing to bear those large frameworks themselves is not something that can be accomplished by computer algorithms, no matter how sophisticated they may be. A justification of that claim will require an analysis of the nature of wisdom and of what it entails that eludes Internet search mechanisms.
It is indeed generally taken for granted that there is a hierarchy in levels of learning, from data to information and opinion, to knowledge and understanding, with wisdom often placed at the top of the sequence. I think there is an important qualification to that positioning: as conventionally understood, wisdom seems to have one foot on the hierarchical ladder and another off to the side. It usually does entail cognition of a sort—i.e., an intellectual apprehension of the true—and also at a level of comprehension above that of “understanding.” In this sense, what I’ll call “cognitive wisdom” does indeed occupy the space at the top rung of the ladder. But wisdom as a whole does not reduce to “cognitive” (or “intellectual” or “speculative”) because it also entails the habit of willing the good in conduct—a crucial aspect separate from any of the ladder’s intellectual steps. The moral component ideally surrounds or envelops the ladder, but experience shows that it may also be lacking even when all of the cognitive steps from data up to the top are present.
This aspect of wisdom consists in the habit of living, or at least striving to live, according to the virtues. A traditional list of these would include prudence, justice, fortitude, and temperance, if not faith, hope, and charity or compassion as well. It is this ethical aspect that often leads observers to note that many people who have little formal education—and who may also live within widely variant religious, agnostic, or atheistic creeds—nevertheless live wisely or sagely,1 while, conversely, many who have great knowledge or understanding of particular subjects are nonetheless foolish in the practical conduct of their own lives.
The possession of wisdom as a whole, however, does not simply reduce to acquiring the habit of acting on moral values such as prudence, justice, or compassion; it also entails an intellectual insight of being able cognitively to discern a reasonably justifiable grasp of the ultimate structure of the “forest” (i.e., the universe and our experience of it) amid the myriad “trees” of data—opposed, at least, to mistaken views assumed by those who are recognizably ignorant, unwise, shortsighted, or foolish or opposed to views that, playing out over time in the laboratory of history, have repeatedly led to disastrous consequences of war, genocide, oppression, or large-scale misery and injustice.
The achievement of wisdom, both cognitive and moral, entails considerations of “what needs to be brought to bear on the question at hand” that cannot be captured by any relevance-ranking keyword searches. The reason is obvious: the same data and information can be framed very differently according to different belief systems. Such systems themselves are variable according to their cognitive assumptions of what is the ultimate source of meaning. A list of only a few candidates for providing that ultimate ground for meaning would include the Biblical God, the Qur’anic Allah, the pantheistic One, the Void (or nonbeing/nonexistence), Marxist dialectical materialism, matter and energy, mathematical elegance, endless cosmic cycles or multiverses, Platonic Forms, Aristotelian form and matter, Darwinian evolutionary success and genetic diffusion, sustainable environmentalism, societal conventions based on power relationships, one’s own self or will, one’s family or tribe or social group, the interaction of yin and yang, sexual fulfillment—and many others.
There seem to be some important considerations that are not entirely arbitrary—while also being not “provable” one way or the other, either—entailed in the acceptance of one ultimate intellectual framework or ordering principle in comparison with another:
Quite obviously no computer “relevance ranking” can bring to bear any of these concerns in patterning which configuration of information is preferable to the alternatives. Disagreements about such matters, however, are not without consequences.
I will not presume to make a case here for any one of the above “ultimates” over the others—they are not all mutually exclusive in any event—but I would point out that the cognitive choice or even tacit assumption of any one of them entails moral value assumptions that are not universally shared by the others. These values in turn serve to frame not just perceptions of what counts to begin with as relevant data and information, but also to contribute in forming the networks of coherence criteria by which the data and information become accepted as adequately integrated knowledge or understanding. The same data and information can be received through multiple different lenses or filters that change, block, or highlight what is perceived as “relevant,” particularly when judgments leading to actions are required. One has only to look at current social debates on gun control, immigration, assisted suicide, or government/corporate responsibilities to see that such problems can never be resolved by information retrievals created by computer algorithms underlying Internet (or other) search boxes. The “relevance” of data and information determined by algorithms working on specified keyword search terms cannot determine the relevant cognitive frameworks or moral values by which the data and information are to be integrated. Any such incapacity is an incapacity to convey wisdom.
There are particular dangers lying in the fact that “information” (as well as opinion) tends to be mistaken as either knowledge or understanding and that the “relevance” of information determined by algorithms can be uncritically taken as “value-neutral” in the sense of somehow floating above partisan concerns and not characterized by distorting filters or selectivity imposed by the algorithms themselves. Bluntly, the concealed proposition would be something like “since the algorithms cannot impose values shaped by any particular worldview framework of meaning, their determination of ‘relevance ranking’ in what they retrieve can simply be taken as ‘objective.’”
Internet searching, however, does impose a framework of its own in determining what is “relevant” to any inquiry, and that frame might best be described as a kaleidoscope view. In a kaleidoscope one sees many individual small pieces of colored glass that are indeed arranged in patterns or larger structures, often of great complexity, but the patterns and structures themselves are quite unstable, and the change in position of even one or two individual pieces can bring about cascades that radically alter the overall view. The analogy is not perfect here, but it is nonetheless readily confirmable that Google, Bing, Yahoo!, or other Internet searches on the same terms can bring about results that are substantially different among themselves, and even within each system the retrievals will vary not only from one day to the next but even from one minute to the next. The turning of the Internet kaleidoscope never stops. One might well ask, then, “Are the first three screens of any Internet retrieval reliably indicative of the truth of the subject?”
The relative stability of the top positioning of Wikipedia articles within algorithmic rankings is especially problematic in conveying the concealed proposition that this source itself is “above the fray” in its objectivity. Granted, there is no better mechanism available on the Internet for providing general overviews of what I have called “the shape of the elephant” of the many subjects it covers, but the Wikipedia’s shaping itself is variable, very frequently unsupported by adequate documentation, and often lacking in comparative perspectives. Further, no encyclopedia-length article is ultimately as intellectually anchored as a scholarly book-length text. Wikipedia is generally good for what it does at the levels of data and information, but it, too, cannot convey the relevance of large worldview perspectives on the subjects it treats of. The danger here is its conveyance of the concealed proposition that such larger perspectives are routinely irrelevant to the information at hand.
There is thus an unarticulated assumption that the avoidance of any particularly specified worldview framework provides intellectual objectivity or moral neutrality, but this very assumption subliminally conveys its own worldview—that the shifting kaleidoscopic arrangement of data and information is the best determinant of meaning that can be appealed to. This perspective itself, however, needs to be compared to the other frames based on quite different assumptions of what is the best or ultimate determinant of meaning.
This is where wisdom comes into play in ways that cannot be reached by computer algorithms: wisdom is the mechanism by which we determine what counts as evidence to begin with according to overarching or ultimate standards of truth, goodness, and beauty. And those standards themselves are not matters of mere opinion, but can be judged by the range and adequacy of which considerations (as only partially suggested by the above list) they take into account as requiring explanation to begin with.
Once more: no computer algorithm can bring to bear any of these large and unavoidable framing assumptions determining the “relevance ranking” of information retrievals.
If nothing else, such considerations should reasonably restrain those of us in the information profession from claiming too much for what we can actually deliver in furthering the goals of either formal education or informal human learning in general. Grandiose assertions that either “all information” or even “all knowledge” can be conveyed electronically (and immediately) stumble badly, to begin with, at the lower levels of the learning hierarchy: our best algorithms cannot distinguish either understanding or knowledge from information or opinion; further, they are oblivious to the choice of one or another ultimate ordering principle, the selection of which has profound consequences in shaping our responses to, and evaluations of, both new situations and new information. Indeed, that choice has a major influence in framing what we judge to be worth attending to right from the start in defining what we accept as knowledge or understanding.
Emerson once wrote, “it is not observed . . . that librarians are wiser men than others.” If I may paraphrase the sage of Concord, “it is not observed that either librarians or information scientists are wiser people than others.” This is particularly true if we regard the goal of our profession as achievable by providing “one-stop” access to “all” information at once—i.e., by breaking down the boundaries of carefully crafted disciplinary “silos” and asserting that “a single search” box backed by algorithmic relevance ranking of keywords is all that we need to provide. Such a belief—and such a practice—would be deadly to the conveyance of the higher levels of learning that our culture requires. If we as a profession wish to promote scholarship at its highest levels, perhaps even those approaching the achievement of wisdom, we cannot rely on either Internet content or search mechanisms alone to do the job for us. We have to provide real alternatives, not only covering important sources that lie in blind spots to Internet searching, but also in also in providing access to those sources in ways that enable recognition of what cannot be specified in advance. Real libraries staffed by knowledgeable librarians and curators are that alternative.