game, in the long run adaptive success attends thosenversions of our partisanship which have the widest, panhuman,nappeal Let us not seek to avoid bias but to widennour bias in favor of the whole human race and beyond.nThis approach especially questions the apparent straightforwardnessnof the notion of political power. Events occur,nand their meaning is rich and complex. The events arenmade up of the actions of men and women; and if theynperformed those actions then tautologically they had thenpower to do so. Do we gain anything by inserting the idea ofnpower? Suppose they didn’t perform the actions; could theynhave? Could we prove it? Power depends on values, andnvalues on the individual and collective imagination.nEven the very methodology of historical research maynhave to be profoundly modified if the new view of history isnto prevail. A perceptive critic of historiography. Gene Wisen{American Historical Explanations) has already pointed outnsome of the fallacies in the accepted model of historicalnresearch, in which “primary sources” (contemporary documentsnand suchlike) are valued more highly than secondarynones (e.g., later interpretations by historians), and they innturn are preferred over tertiary discussions and revisions ofnhistorical interpretations. He rightly declares that the onlyntrue primary source is the actual experience of the participantsnin historic events. All other sources are secondary andnpartial filters; but the nature of the filter itself is part ofnhistory, and perhaps a more crucial part than we think.nR.G. Collingwood said: All history is the history of ideas.nOr we might put it this way: history (small h) is Historyn(capital H).nPerhaps we could go even further and remark that muchncontemporary documentation may paradoxically be worsenevidence than later judgments by hindsight. Consider annanalogy: the information-processing system of the humanneye and visual cortex. If “primary” evidence is better thannsecondary, then the best visual data about what is happeningnin the “outside world” is the raw firing of the retinalnneurons—the work of the optic nerve and visual cortexnwould amount to nothing more than a heap of opinion andnsophistry. The military technologists who wanted to designnan eye by which homing missiles might recognize their preyntook this position at first but were soon proved wrong. Theirnattempt to hook up a TV camera to a simple computernprogrammed with pictures of enemy tanks simply did notnwork. The raw data did not add up to tanks but to a riot ofnshadows, colors, changes in reflectivity and albedo, geometricndistortions by perspective, and confusing shiftsnwhich could not distinguish between subjective motion,nmotion of the object, changes in the object, and changes innlight and shadow. To make a catalog of all the appearances antank might take on would require a memory as big as thenuniverse, which, moreover, could easily mistake somethingnelse for a tank. Later work by Artificial Intelligence researchersnlike Marvin Minsky and vision scientists likenEdwin Land and David Marr revealed the astonishingnhierarchy of ganged or independent servomechanisms thatnmakes up vision, and its radical dependence on priornexpectations, needs, or questions to make any sense of thenworld at all. In other words, the secondary and tertiarynsources are much more reliable than the primary ones, tonuse the old classification.nFurther, of course, the physical world is, as we know nownfrom quantum theory, constituted by the cooperation ofnsentient beings. Thus tanks don’t look like tanks and can’tnact as tanks without being interpreted as tanks; their tanknexistence derives from the secondary and tertiary data. Andnof course this analogy is more than an analogy but a furtherndeconstruction even ofWise’s notion of “experience.” If thenold hierarchy of primary and secondary is so problematicnfor visual experience itself, think what it is for the process ofnhistorical evidence-gathering. What history really is is ourninterpretation of it. We are all, in the deepest sense ofnexperience, contemporary experieneers of those historicalnevents, quite as much as the astronomers are contemporarynexperieneers of the supernova that exploded 160 millionnyears ago and whose light has just now hit their telescopes.nWe can with relief relinquish the positivism of the oldnmainstream historical research; after all, the physicistsnabandoned it 50 years ago.nWe can abandon, too, that kind of history which assumesnalways that there is a true, hidden version of events, ofnwhich the apparent surface is a hypocritical cover-up. Ofncourse, hidden motives and interests do govern what goesnon; but everybody has such motives, and everybody is anplayer in the game, a partner in the feedback. If thenpopulation is deceived, it may be because it wants to bendeceived and, if the deceiver is unveiled, will get itself anmore competent magician and a more satisfying illusion.nAnd such collective creations, in this contingent world ofnours, verily constitute reality. The surface of history is thenreality, and the new historiography will treat it as annexpressive, meaningful, but inexhaustible artistic object,nand not seek only to reveal an inner truth that derives fromninitial conditions or a conspiratorial deus ex machina.nThis means that the capacity to recognize beauty, thenaesthetic sense, is the primary cognitive skill of the historiannor sociologist. It is by beauty that we intuit the order of thenreflexive process of human history. On the small, tribalnscale, the need for this essential function may well havenbeen one of the principal selective pressures that led usntoward our extraordinary inherited talents at storytellingnand the interpretation of narrative. History should benrefounded on story, not the other way around. Indeed,nmuch of the most exciting new work in sociology—forninstance, the work of Thomas Scheff—is beginning to takenthis direction and to break free from its positivistic traditions;ncultural anthropology, with its participant-observernmethods and its tradition of listening to native informants,ndid not have very far to go.nIf the new definition of historical understanding is thatnwe understand what we are able to model, and if our newndefinition of evidence is experience, how do we model andnexperience history?nI would suggest a number of ways, some of which arenalready being used unsystematically and without theoreticalnjustification, but very effectively nonetheless. Almost all ofnthem involve a greater or lesser degree of collapse betweennthe activities of teaching, learning, and research—whichnwould be no bad thing for our academics.nFirst, historical reenactments. Popular culture, as often,nis ahead of the academies. The reenactment of Civil Warnbattles and the battles of Texan independence has become annnDECEMBER 19871 37n