In lieu of an abstract, here is a brief excerpt of the content:

September 2003 ยท Historically Speaking 37 Predictions and Analytical History Bertrand M. Roehner When you visit your doctor because (for instance) you feel stomach pains, she will askyou a number of questions. Do you feel feverish? Do you feel pain immediately afteryourmeals, and so on? Then your doctor will try to establish a link between the description framed by your answers and one ofthe recognized patterns of symptoms thathave been recorded in die corpus of medical science. Depending on the symptoms she maydiagnose aviral infection, an ulcer, orwhathave you. Three points need to be emphasized: * There is never absolute certainty about the diagnosis. Not infrequently, starting from the same set of symptoms, two doctors would make different diagnoses. Each patient in some wayis unique, and to some extent each illness is unique, too. However, both patients and diseases clearly share common features. * Only new evidence can help to decide which diagnosis is correct. * The ultimate test for deciding which interpretation is correct is to see which treatment is able to cure the patient. Of what relevance is this to Pattern and Repertoire} Formesakeofillustration, consider theexampleofpricepeaksinrealestate. Today there is a hot debate in most developed countries regarding die future ofreal estate markets . Some economists claim that demand is strongandwillremainso; others contend that theremaybeasmall correctioninthenexttwo years. But surprisingly few researchers have tried to take a close lookat formerprice peaks to betterunderstandwhatmaybe in store. Yet readers in Boston, NewYork, or San Francisco mayrememberthatthirteenyears agothebust ofa housing bubble brought about financial troubles formanysavingsinstitutions. Willdie samescenariorepeatitselfin the five orsixyears to come? Needless to say, the economic environment is not exactly the same in 2003 as it was in 1990; but, as we have seen, neitherare all symptoms identical for two patients. Yet this does notpreventmedical treatment from being successful. The onlysensible attitude is to say: "Well, let's examine each prediction and see whichoneproves correctinmenextfiveyears." One can find a prediction ofthat kind in the June 2003 issue ofMoney: real estate prices in SanFrancisco should climb by 10.7% between June 2003 andJune 2004. 1 have published another prediction according to which prices in San Francisco should fall by 10% duringthe same period.1 This prediction is based on a careful examination ofall previous housing price peaks in the U.S. as well as in London, Paris, and Sydney. In the fall of2004whenprice data forJune will become available, the readers ofHistorically Speaking will be able to see for themselves which approach works best As a matter of fact, a similar test can be mentioned in this connection. In May 2000, shortly after prices on the NASDAQ stock market began to fall, I made a prediction for the evolution of the NASDAQ Composite Index ovef the nextsixyears.2 So far die prediction has proved fairly accurate. Needless to say, itwas based on die observation ofprevious stock market crashes: die crash of die Paris stock market in 1881; die New York crash ofOctober 1929; and die Tokyo crash in late 1989. In each case diere were ofcourse substantial differences in die overall business situation; nevertheless, die economic frenzy and imbalance were verymuch die same. Itis precisely die fact mat diese price peaks were so huge which made prediction possible. The phenomenon is very similar in medicine: if John and Harryhaveamild flu, dieymaypresent fairly different symptoms because their organisms may react in different ways, especially ifJohn is twentywhereas Harry is seventy . However, for more severe infections dieir reactions might be more similar. Whenwe establish a parallel between two historical episodes, for instance die meeting of die Estates General in France in 1356 and 1789 (Pattern andRepertoire, 108), is such a parallel completely arbitrary? The answer is that it is no more (and no less) arbitrary than the link physicians establish between die sets of symptoms described by patients John and Harry. Unfortunately, in the case ofthe meetings ofthe Estates General there is no way to prove that the modular way oflooking at history is more insightful than others; since the Estates General no longer exists there can be no testable prediction in this respect. Failing diat, die assessmentwill relyon howconvincingorpersuasive the comparison sounds; these are subjective notions, however. The same comparisonmightwell be found enlightening by some readers but unconvincing by...

pdf

Share