130
148 SOCIAL RESEARCH
So far, however, systems theory has used only very simple, chrono- logical notions of time and future, conceiving of the future simply as the state of the system at a later time.
148 SOCIAL RESEARCH
So far, however, systems theory has used only very simple, chrono- logical notions of time and future, conceiving of the future simply as the state of the system at a later time.
The-future-cannot-begin-Niklas-Luhmann
of course.
the main idea of George Herbert Mead.
Mead himself uses the formulation "past pasts" in the sense of pasts of past presents.
Cf.
Mead.
The
PhilosoPhy 01 the Present, p. 7.
81 For the medieval discussion de futuris cont;ngePltibus and its importance for
church policy. see Thomas Aquinas. In I. Per; Hermeneias lect. XIII, XIV: Qua-
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES 141
discussed with respect to the limits of logic and human cognition in its application to future events-and not as the technique of defuturizing the future by the binary code of logic.
Whereas the ancients started with generalizations of their every- day world by means of cosmological and theological assumptions and thought not of "the" future but of coming events and the possi-
bility of their privative negation. s2 we experience our future as a generalized horizon of surplus possibilities that have to be reduced as we approach them. We can think of degrees of openness and call /utur;zat;on increasing and de/uturizat;on decreasing the openness of a present future. Defuturization may lead to the limiting condition where the present future merges with the fu- ture presents and only one future is possible. Actually. the struc- ture of our society prevents defuturization from going this far. But there are techniques of deflIturization which react exactly to
this condition. Leon Brunschvicg has drawn our attention to the fact that the statistic calculus defuturizes the future without identifying it with only one chain of events. ss And indeed. the new interest in chance. games of hazard. and statistics coming up
in the seventeenth and eighteenth centuries corresponds closely to an emerging interest in the future and to the idea that it may be a rational and even a secure strategy to prefer the insecure over the secure. S4 There are ways to make use of the future without beginning it and without reducing it to one chain of datable future presents.
estiones disputatae de Veritate q. 11, art. 12; Summa Theologiae I q. 14 art. 15; William Ockham, Tractatus de praedestinatione et de praescientia Dei et de futurls contingentibus, edited by Philotheus Boehner (St. Bonaventure, N. Y. : Francisc:an Institute, St. Bonaventure College, 1945); Leon Baudry, ed. , La Querelle des futurs contingents (Louvain 1465-1475) (Paris: J. Vrin, 1950).
a8 Cf. Paulu8 Engelhardt, "Der Mensch und seine Zukunft: Zur Frage nach dem Menschen bei Thomas von Aquin," in Festchrift fur Max Muller (Freiburg- Munchen, 1966), pp. 852-874.
aa Leon Brunschvicg, L'experience humaine et la causalite physique (Paris: Alcan, 1949), p. 855?
. . Cf. Ernest Coumet, "La Th~orie du Hasard est-eUe nee par Hasard? ," Annales: Economies, Sodetes, Civilisations 25 (1970): 574-598.
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
142 SOCIAL RESEARCH
Temporal Integration Redefined: Technology and Utopian Schemes
By nowl we are advanced far enough to redefine the problem of temporal integration. One possible interpretation would be that te~poral integration is achieved by changing wishful thinking and fanciful perspectives into more realistic ones, adapting to the out- come of the past so far as it has structured the present. 81S This view evaluates realism as maturity. But why so? If lower-class children abandon certain educational and occupational aspirations, this may be so much the better for them. It would be rational, how- ever, only insofar as reality itself is rational. T o identify temporal integration with realistic orientation presupposes a perfect world -realitas sive perfectio. This is a well known traditional premise, but it does not differentiate time and reality far enough to use temporal integration as a means to control-not necessarily to change-reality.
There have been societies which had to use reality as rationality control. Our society, however, has to use rationality as reality control. Its structure and its environment are too complex for adaptive procedures,86 and there is not enough time available for adjustment. Under the condition of high complexity, time be- comes scarce. Time has to be substituted for reality as the pre- dominant dimension while future obtrudes itself as the predom- inant horizon. Such a society will need forms and procedures of
temporal integration which, above all, combine the present future and future presents and consider the past only as th. e set of facts which we are no longer able to prevent from existing or becoming.
The prevailing conception of the present future seems to be a utopian one 8T with an optimistic or a pessimistic overtone. The
1111 See, for example, Cottle andKlineberg, The Present of Things Future, pp. 70 If.
lie Russel L. Ackolf and Fred E. Emery, On Purposeful Systems (London and Chi? cago: Aldine, 1972), esp. pp. 80 If, pursue a similar intention by distinguishing goal. seeking and purposeful systems.
liT In one important sense the reference to "utopias" is misleading here because originally the literary device of a utopia was invented Just because critics were 9101
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES 143
future serves as a projection screen for hopes and fears. Its uto- pian formulation warrants rational behavior toward different (predictable and unpredictable) future presents, at least in the form of coherent negation. The future is expected to bring about the communist society or the ecological disaster, emancipation from domination or l'homme integrale discussed by Sartre and Merleau-Ponty. 38 This is the future that cannot begin. It re- mains a present future and at least an infallible sign of the pres- ence of critics. It moves away if we try to approach it. It does not
vanish, however, as long as the structural conditions of the present society endure, but it may resettle with new symbols and meanings, if the old ones are worn out by disappointments and new experi- ences. Our recent experiences seem to show that these utopian futures speed up their change and may change so quickly that they never will have a chance to be tested and to get confirmation in a
present.
Technologies, on the other hand, orient themselves to future
presents. They transform them into a string of anticipated pres- ents. They postulate and anticipate causal or stochastic links be- tween future events in order to incorporate them into the present present. This implies two important reductions of complexity. The first transforms the character of events which are emerging recombinations of independent contingencies into a carrier func- tion of the process of determination. The second brings into re- lief a sequential pattern, a chain of interconnected events; it se- quentializes complexity by abstracting more or less from inter- fering processes. 39 A future defuturized by technology can be
able to use the future of their own society as projection screen. The turning point can be dated exactly: in 1768 Mercier began to write his l'An deux mille quatre cent
quarante.
88 A comprehensive presentation of such imaginary approaches to future is Fred L. Polak, The Image of the Future, 2 vols. (New York: Oceana Publications, 1961). However, it does not pay enough attention to the historical variability of time itself. Cf. also Wendell Bell and James A. Mau, "Images of the Future: Theory and Re? search Strategies:' in Bell and Mau, eds. , The Sociology of the Future: Theory, Cases, andAnnotatedBibliography(NewYork:RussellSageFoundation,1971),pp. ~.
89 A harsh criticism of the technocratic conception of time has been formulated by
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
144 SOCIAL RESEARCH
used as a feigned present from which we choose our present present to make it a possible past for future presents. To justify the choice and, more important, to justify this whole procedure of technical defuturization we use values. Values, then, have the function of guaranteeing the quality of present choice in spite of technical defuturization. Any refinement, however, of techno-
logical forecasting and control will make future presents so much more surprising, because it multiplies defeasible assumptions about the present future. It requires, therefore, in its present, corre- sponding mechanisms of coping with surprise: learning potential, planned redundancie~, and the generalized ability to substitute functional equivalents.
Technology and utopian schemes are, of course, very different approaches to the future. Their difference suggests options and polemical behavior. Many ideological discussions and political confrontations of our day draw their resources from this bifurca- tion. If you embark on the vessel named Utopia, you will be- come highly critical in respect to technology, and rightly so, even if you are prepared to use technology to get your vessel off the shores. If, on the other hand, you set out to improve technology you may get annoyed, and again rightly so, with people who use the future as a substitute for reality and interfere with your work without contributing to it. Each side tries to totalize its own perspective on the future and suppress the other. 40 But the totality
Herbert G. Reid, "The Politics of Time: Conflicting Philosophical Perspectives and Trends," The Human Context 4 (1972): 456-483; "American Social Science in the Politics of Time and the Crisis of Technocorporate Society: Toward a Critical Phe-
nomenology," Politics and Society 3 (1973): 207-243.
40 This is, of course, what Habermas has in mind when he unveils the use of tech-
nology and systems theory as ideology. Cf. JUrgen Habermas, Tec/mik und W;ssen- schaft als "Ideologie" (Frankfurt: Suhrkamp, 1968); Jtirgen Habermas and Niklas Luhmann, Theorie der Gesellscllaft oder Sozialtechnologie-Was leistet die System- forschung1 (Frankfurt: Suhrkamp, 1971). See also Robert Boguslaw, Tile New Utopians: A Study of System Design and Social Change (Englewood Cliffs, N. J. : Prentice-Hall, 1965); Joseph Bensman and Robert Lilienfeld, Craft and Con- sciousness: Occupational T echnique and the Development of W orld Images (New York: Wiley, 1973), pp. 282 ff; Robert Lilienfeld, "Systems Theory as an Ideology," Social Research 42 (Winter 1975): 637~60.
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES 145
is the difference itself: the difference of the present future and future presents. This difference itself is a historical fact, pro- duced and reproduced by the structure of our society. W e cannot avoid it or circumvent it as long as we continue to live in this highly complex society. But this does not mean that we have to pursue these pointless polemics.
Still, critical discussion and polemics have the important ad- vantage of being present behavior. Any attempt to replace them by posing the problem of temporal integration would defer the solution of this problem into the future and would, thereby, slide off into either utopian or technical channels. Again, the prob- lem of temporal integration, too, would become either a utopian or a technical problem and, thus, perpetuate itself.
An open and indeterminate future seems to suggest a shift from cognition to action, as Marx would have it, or today from pre- dicting to creating the future. 41 This sounds like: If you can- not see, you have to actl But both, prediction and action, have their utopian and their technical aspects. Substituting the one for the other does not solve the problem of temporal integration. The complex society of our day has to use both ways for reducing the complexity of its future; it has rather to sequentialize predic- tions and actions into complex self-referential patterns. There is no problem of choice between prediction and action, but there may be a problem of social and structural limitations for the com- bination of predictions and actions.
Social Communication as a Nontemporal Extension of Time
It should be clear by now that we can expect temporal integra- tion and, for that matter, integration of utopian schemes and tech- nology only as a present performance. Therefore, older societies which thought of themselves as living in an enduring or even
U So Bettina J. Huber, "Some Thoughts on Creating the Future," Sociological In- quiry 44 (1974): 29-39.
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
146 SOCIAL RESEARCH
eternal present did not experience our problem. Only in modern times. and only by shortening the time span of the present. does the problem of perseverance. or conservatio, get its actuality. 42 and only then do utopian schemes and technology diverge. By re-
structuring time in the last 200 years. the present has become specialized in the function of temporal integration; however. it
does not have enough time to do this job.
It is at this point that we can grasp the importance of the
theoretical contributions of George Herbert Mead 48 and Alfred Schutz44 concerning the interrelations between temporal and social experience. Both authors were aware of the fact that social communication defines the present lor the actors (because it com- mits the actors to the premise of simultaneity) and provides in addition the chance lor a nontemporal extension 01 time. "The field of mind. " in the words of Mead. "is the temporal extension of the environment of the organism. " and the mechanisms which accomplish this are social ones. 41i But then. the environment of systems can be also used as a nontemporal extension of time.
Other persons are socially relevant only insofar as they present. in communication. different pasts and/or different futures. They transform in a highly selective way distant temporal relevances into present social ones. And it is this selectivity that can be sub- mitted to social control-for example. by the twin mechanisms of trust and distrust. 46 This nontemporal extension of time by com-
munication constitutes time horizons for selective behavior-that is. a past that can never be reproduced because it is too complex and a future that cannot begin. And it is again this temporal com-
42 Cf. Hans Blumenberg. Selbsterhaltung und Beharrung: Zur Konstitution der neu%eitlichen Rationalitiit (Abhandlungen der Akademie der Wissenschaften und der Literature in Mainz. Wiesbaden, 1970).
48 Mead, The Philosophy 0 / the Present.
44 See above all Alfred Schutz, Der sinnha/te Au/bau der so%ialen Welt (Vienna: J. Springer, 1982).
411 Mead, The Philosophy 0/ the Present, p. 25.
40 For a more extensive treatment, see Niklas Luhmann, P'ertrauen: Ein Mech-
anismus der Reduktion so%ialer Komplexiliit, 2nd ed. (Stuttgart: F. Enke, 1978).
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES
147
plexity that makes selectivity necessary for meaningful behavior and communication.
These considerations bring us back to the roots of evolutionary interdependencies between social and temporal structures. Since this can be regarded as achieved knowledge, we cannot afford to
fall back on much simpler notions of the future as most social fore- casting does. The conception of interdependency, however, is in itself too vague and indeterminate to serve as a framework for further analysis. Neither Mead nor Schutz had adequate suc- cessors. The next step, indeed, is a difficult one. It requires the conceptualization of limitations and of gains that might result from novel combinations.
In view of the facts our society has produced in its bourgeois phase we should be able to calculate the limits of the meaningful extension of time; we should know the social correlates of a high differentiation of temporal horizons; we should be able to antici- pate a change in temporal structures as a consequence of social change-for example, as a consequence of an eventual decline of the monetary mechanism; we should be able to estimate the degree of heterogeneity of temporal structures we can tolerate in different subsystems of our society; we should know how the shrinking temporal horizons of families affect the economy, and how we can avoid the well known negative impact which the time perspectives of a growing economy have on the political system; 47 and, last but
not least, we should know what is implied if we rely on clocks and dates to integrate the different time perspectives of different sectors of the society and what dysfunctional consequences we have to expect if we use chronology to fulfill this important function.
It is sure that we cannot reduce this set of complex questions, involving the future, to a single one: how to begin the future. It is difficult to see how we could proceed in elaborating these questions or even answering them. Systems theory seems to be the only conceptual framework which has sufficient complexity.
61 For a classical statement, see Alexis de Tocqueville, L'Ancien regime et la revolu- tion. 5th ed. (Paris, 1866).
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
?
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p.
130
148 SOCIAL RESEARCH
So far, however, systems theory has used only very simple, chrono- logical notions of time and future, conceiving of the future simply as the state of the system at a later time. 48 Only environment, but not time, is recognized as a set of possible restraints on system states. Abstracting from time is, of course, quite legitimate as a scientific procedure; but then we must refrain from using tempo- ral notions in presenting the results.
In comparison with the conceptual elaboration of problems of time, systems theory 'is much more advanced in its conceptual com- plexity. It is the theory of time that is lagging behind, not the theory of systems. Not only social science but also the theory of history suffers from this deficiency. If the theory of time could be advanced, there Inight appear highly suggestive possibilities of research in correlations between system structures and temporal
structures.
The theory of time has to transform its vague idea of "every-
thing is possible in the long run," based on a chronological con- ception of time, into a concept of temporal structures with limited possibilities of change. It is a prerequisite of correlations that both variables are reduced contingencies in the sense that they cannot assume any shape whatever. We have, therefore, to look for time-inherent restrictions of possible correlations (substituting this for older notions of the substance or essence of time) before we set out to establish correlations between system structures and temporal structures. These time-inherent restrictions are, never- theless, results of sociocultural evolution and not a priori as- sumptions about the nature of the world or conditions of cogni- tion.
If we conceive of time as the relation between (more or less differentiated) temporal horizons and if we use a conceptual lan- guage that allows for iterative modalizations (present future, fu-
ture presents, future of past presents, etc. ) and define the function of the present and the function of chronology in these terms, we
48 See as a rather typical example Ervin Laszlo, A Strategy for tile Future: The Systems Approach to World Order (New York: Braziller, 1974).
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES 149
may have a sufficient base to start this kind of research. But we have to remain aware of the fact that a commitment to these con- ceptualizations is a commitment to "modern times. " Older so- cieties did not produce such an elaborated framework, and they did not need it to understand themselves. They lived, for struc- tural reasons we may be able to explain, within a less differentiated time.
The Future of Systems
Social systems are nontemporal extensions of tilne. They make the time horizons of other actors available within one contempo- rary present. This requires for social systems a double relation to time: a sequential one conceivable as process or as action in terms of means and ends, and a structural one conceivable as the differ- ence between system and environment. With respect to time, the difference of system and environment means that no complex sys- tem can rely exclusively on point-to-point relations to its environ- ment-that is, on instantaneous adjustment by immediate experi- ence and immediate reaction. 49 It needs time for its own opera- tions. This presupposes that under normal conditions no single event will change the whole system at once. Changing everything at once amounts to destruction. In other words: There is no con- ceivable state of a complex system which could be achieved by changing everything at once. The structural technique by which a system avoids this condition of changing everything at once is differentiation-or more exactly: a matching of internal and ex- ternal differentiation. lSo It is only at this rather taxing theoretical level of the relation between the relations of system/environment
49 Cf. Talcott Parsons, "Some Problems of General Theory in Sociology:' in John C. McKinney and Edward A. Tiryakian, eds. , Theoretical Sociology: Perspectives and Developments (New York: Appleton? Century? Crofts, 1970), pp. 27-68.
110 Cf. W. Ross Ashby, Design for a Drain (New York: WHey, 1952). Cf. also Uriel G. Foa, Terence R. Mitchell, and Fred E. Ficdlcr, "Differentiation Matching," Be- hav;oral Science 16 (1971): 130-142.
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
150 SOCIAL RESEARCH
and structure/process that we are able to locate our problem. Systems, then, in relation to their environment, depend for tem- poral reasons on a differentiation of structure and process. The
time perspective of modern society, on the other hand, project,s the difference of the present future and future presents. Both distinctions, worked out in very different intellectual traditions, seem to converge. If this is true, we can bring together systems
theory and phenomenological research.
In fact, the process of continuing communication in social sys-
tems under the condition of contemporaneity is the prospect of sequential social presents that will constitute forever new futures and new pasts. They are and will remain presents because they require a simultaneous integration of the perspectives of different actors. Structure, on the other hand, establishes for our society an open future in the sense that it provides for the selectivity of future presents.
Stated in more concrete terms, structure makes it possible and even necessary to postpone choices and to use the present future as a kind of storehouse for decisions to be made later. At the same time, the present system operates on the premise of continuing its processes. As a system it reproduces its present step by step. This sequentializing of presents, however, is meaningful only as a chain of choices, not as a chain of facts. The process of communication has its effect in producing and reproducing choice situations.
Going further, we have to break up this general notion of post- ponement of choices and have to distinguish two essentially differ- ent forms: (I) deferment of gratification and (2) deferment of ne-
gation. Both have their functional and institutional correlates. Deferment of gratification is a main prerequisite for the economic system as a condition for capital investment. Deferment of nega- tion is a main prerequisite of the political system as a condition of trust in political power. Both require institutional support, both require a present future for their present motivation. Both re- quire a working integration of utopian schemes and technology
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES 151
and a kind of security base for trust. III Both would not survive a considerable shrinkage of time horizons. Both may be endan- gered by too high a fluctuation rate of utopian schemes and tech- nological innovations. And, last but not least, do we not take too much for granted that it is and will remain possible, in spite of changing structural conditions, to separate deferment of gratifica- tion and deferment of negation and to avoid spill-over effects? Or will a refusal to defer gratifications any longer amount to a re-
fusal to defer negations; and finally, will the shrinking of time horizons in the economy endanger trust in politics, political ideol- ogies, value schemes, etc?
All of these questions pertain to what we have come to call bilrgerliche Gesellschaft and relate to the continuity or discon- tinuity of its structures under changing conditions. The bilrger- liche Gesellschaft has been a revolutionary society with a strong structural emphasis on time and corresponding simplifications of social and environmental relations. The principle of its future was simply the denial of its past 112 by the antistructural postulate of equality. lls The self-conception of this society"in its bourgeois variant did rely heavily on time-using and time-binding mecha- nisms like money and legal procedure. By now, we are aware of
111 For the function of security bases in relation to generalized media. of communi- cation. see Talcott Parsons, "On the Concept of Power" ~md "On the Concept of InOuence. " in his Sociological Theory and Modern Society (New York: Free Press, 1967). pp. 297-354. 355-382. Furthermore, Niklas Luhmann. "Symbiotische Mech- anismen," in Otthein Rammstedt, ed. , Gewaltverhiiltnisse und die Ohnmacht der Kritik (Frankfurt: Suhrkamp, 1974), pp. 107-131.
112 Cf. Joachim Ritter, Hegel und die franzosische Revolution (Koln: W est? deutscher V erlag, 1957).
113 A well known statement is Antoine de Condorcet, Esquisse d'un tableau his- torique des progres de l'esprit humain (1794). For the continuing impact of this idea and for empirical correlations between future orientation and emphasis on equality. see James A. Mau. Social Change and Images of the Future: A Study of the Pursuit of Progress in Jamaica (Cambridge: Schenkman. 1968). Since equality implies freedom and freedom implies inequality. the postulate of equality cannot refer to reality. but only to time. Its only function is to deny the relevance of the past--e. g. , the relevance of biographies and ascribed status for the access to educa-
tion (equality of opportunity) or to political elections.
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
152 SOCIAL RESEARCH
highly complex operating conditions and of the narrow limits of effectiveness these mechanisms are subject to. In its Marxist or dialectical variant, the theory of society has to build its concept of future on negations of the present; but there is much more to negate in our present society than dialecticians could ever use for constructing or even bringing about one and only one de- sirable future: They have to focus on one central problem, thus overstating centralization, and to discount complexity in order to design a strictly linear theory which can be used to reconstruct or even to change the "process of history. "
There are many reasons, then, to suspect that the burgerliche Gesellschaft went very far in temporalizing reality and that the twin conceptions of bourgeois and Marxist theory were based on this. common presupposition. This does not decide the question whether this is a temporary distortion characteristic of the period of transition into a new type of world society, or whether this reflects lasting prerequisites of highly complex societies and/or an acceleration of the evolutionary process without parallels in pre-
vious history. We are certainly not prepared to decide this ques- tion without further research on the conceptual as well as on the empirical level. But we have the intellectual resources to go be- yond the boring controversies of Marxist versus- bourgeois or
utopian versus technocratic theory, and the starting positions are available for working out a systems theory of society which recog- nizes the fact that the future cannot begin and which compensates by the higher complexity of its conception of time for what might appear as a loss of future.
-I am indebted to S. Ho)mes. S. Seldman. and A. J. Vidich for comments on an earlier draft of this paper.
PhilosoPhy 01 the Present, p. 7.
81 For the medieval discussion de futuris cont;ngePltibus and its importance for
church policy. see Thomas Aquinas. In I. Per; Hermeneias lect. XIII, XIV: Qua-
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES 141
discussed with respect to the limits of logic and human cognition in its application to future events-and not as the technique of defuturizing the future by the binary code of logic.
Whereas the ancients started with generalizations of their every- day world by means of cosmological and theological assumptions and thought not of "the" future but of coming events and the possi-
bility of their privative negation. s2 we experience our future as a generalized horizon of surplus possibilities that have to be reduced as we approach them. We can think of degrees of openness and call /utur;zat;on increasing and de/uturizat;on decreasing the openness of a present future. Defuturization may lead to the limiting condition where the present future merges with the fu- ture presents and only one future is possible. Actually. the struc- ture of our society prevents defuturization from going this far. But there are techniques of deflIturization which react exactly to
this condition. Leon Brunschvicg has drawn our attention to the fact that the statistic calculus defuturizes the future without identifying it with only one chain of events. ss And indeed. the new interest in chance. games of hazard. and statistics coming up
in the seventeenth and eighteenth centuries corresponds closely to an emerging interest in the future and to the idea that it may be a rational and even a secure strategy to prefer the insecure over the secure. S4 There are ways to make use of the future without beginning it and without reducing it to one chain of datable future presents.
estiones disputatae de Veritate q. 11, art. 12; Summa Theologiae I q. 14 art. 15; William Ockham, Tractatus de praedestinatione et de praescientia Dei et de futurls contingentibus, edited by Philotheus Boehner (St. Bonaventure, N. Y. : Francisc:an Institute, St. Bonaventure College, 1945); Leon Baudry, ed. , La Querelle des futurs contingents (Louvain 1465-1475) (Paris: J. Vrin, 1950).
a8 Cf. Paulu8 Engelhardt, "Der Mensch und seine Zukunft: Zur Frage nach dem Menschen bei Thomas von Aquin," in Festchrift fur Max Muller (Freiburg- Munchen, 1966), pp. 852-874.
aa Leon Brunschvicg, L'experience humaine et la causalite physique (Paris: Alcan, 1949), p. 855?
. . Cf. Ernest Coumet, "La Th~orie du Hasard est-eUe nee par Hasard? ," Annales: Economies, Sodetes, Civilisations 25 (1970): 574-598.
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
142 SOCIAL RESEARCH
Temporal Integration Redefined: Technology and Utopian Schemes
By nowl we are advanced far enough to redefine the problem of temporal integration. One possible interpretation would be that te~poral integration is achieved by changing wishful thinking and fanciful perspectives into more realistic ones, adapting to the out- come of the past so far as it has structured the present. 81S This view evaluates realism as maturity. But why so? If lower-class children abandon certain educational and occupational aspirations, this may be so much the better for them. It would be rational, how- ever, only insofar as reality itself is rational. T o identify temporal integration with realistic orientation presupposes a perfect world -realitas sive perfectio. This is a well known traditional premise, but it does not differentiate time and reality far enough to use temporal integration as a means to control-not necessarily to change-reality.
There have been societies which had to use reality as rationality control. Our society, however, has to use rationality as reality control. Its structure and its environment are too complex for adaptive procedures,86 and there is not enough time available for adjustment. Under the condition of high complexity, time be- comes scarce. Time has to be substituted for reality as the pre- dominant dimension while future obtrudes itself as the predom- inant horizon. Such a society will need forms and procedures of
temporal integration which, above all, combine the present future and future presents and consider the past only as th. e set of facts which we are no longer able to prevent from existing or becoming.
The prevailing conception of the present future seems to be a utopian one 8T with an optimistic or a pessimistic overtone. The
1111 See, for example, Cottle andKlineberg, The Present of Things Future, pp. 70 If.
lie Russel L. Ackolf and Fred E. Emery, On Purposeful Systems (London and Chi? cago: Aldine, 1972), esp. pp. 80 If, pursue a similar intention by distinguishing goal. seeking and purposeful systems.
liT In one important sense the reference to "utopias" is misleading here because originally the literary device of a utopia was invented Just because critics were 9101
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES 143
future serves as a projection screen for hopes and fears. Its uto- pian formulation warrants rational behavior toward different (predictable and unpredictable) future presents, at least in the form of coherent negation. The future is expected to bring about the communist society or the ecological disaster, emancipation from domination or l'homme integrale discussed by Sartre and Merleau-Ponty. 38 This is the future that cannot begin. It re- mains a present future and at least an infallible sign of the pres- ence of critics. It moves away if we try to approach it. It does not
vanish, however, as long as the structural conditions of the present society endure, but it may resettle with new symbols and meanings, if the old ones are worn out by disappointments and new experi- ences. Our recent experiences seem to show that these utopian futures speed up their change and may change so quickly that they never will have a chance to be tested and to get confirmation in a
present.
Technologies, on the other hand, orient themselves to future
presents. They transform them into a string of anticipated pres- ents. They postulate and anticipate causal or stochastic links be- tween future events in order to incorporate them into the present present. This implies two important reductions of complexity. The first transforms the character of events which are emerging recombinations of independent contingencies into a carrier func- tion of the process of determination. The second brings into re- lief a sequential pattern, a chain of interconnected events; it se- quentializes complexity by abstracting more or less from inter- fering processes. 39 A future defuturized by technology can be
able to use the future of their own society as projection screen. The turning point can be dated exactly: in 1768 Mercier began to write his l'An deux mille quatre cent
quarante.
88 A comprehensive presentation of such imaginary approaches to future is Fred L. Polak, The Image of the Future, 2 vols. (New York: Oceana Publications, 1961). However, it does not pay enough attention to the historical variability of time itself. Cf. also Wendell Bell and James A. Mau, "Images of the Future: Theory and Re? search Strategies:' in Bell and Mau, eds. , The Sociology of the Future: Theory, Cases, andAnnotatedBibliography(NewYork:RussellSageFoundation,1971),pp. ~.
89 A harsh criticism of the technocratic conception of time has been formulated by
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
144 SOCIAL RESEARCH
used as a feigned present from which we choose our present present to make it a possible past for future presents. To justify the choice and, more important, to justify this whole procedure of technical defuturization we use values. Values, then, have the function of guaranteeing the quality of present choice in spite of technical defuturization. Any refinement, however, of techno-
logical forecasting and control will make future presents so much more surprising, because it multiplies defeasible assumptions about the present future. It requires, therefore, in its present, corre- sponding mechanisms of coping with surprise: learning potential, planned redundancie~, and the generalized ability to substitute functional equivalents.
Technology and utopian schemes are, of course, very different approaches to the future. Their difference suggests options and polemical behavior. Many ideological discussions and political confrontations of our day draw their resources from this bifurca- tion. If you embark on the vessel named Utopia, you will be- come highly critical in respect to technology, and rightly so, even if you are prepared to use technology to get your vessel off the shores. If, on the other hand, you set out to improve technology you may get annoyed, and again rightly so, with people who use the future as a substitute for reality and interfere with your work without contributing to it. Each side tries to totalize its own perspective on the future and suppress the other. 40 But the totality
Herbert G. Reid, "The Politics of Time: Conflicting Philosophical Perspectives and Trends," The Human Context 4 (1972): 456-483; "American Social Science in the Politics of Time and the Crisis of Technocorporate Society: Toward a Critical Phe-
nomenology," Politics and Society 3 (1973): 207-243.
40 This is, of course, what Habermas has in mind when he unveils the use of tech-
nology and systems theory as ideology. Cf. JUrgen Habermas, Tec/mik und W;ssen- schaft als "Ideologie" (Frankfurt: Suhrkamp, 1968); Jtirgen Habermas and Niklas Luhmann, Theorie der Gesellscllaft oder Sozialtechnologie-Was leistet die System- forschung1 (Frankfurt: Suhrkamp, 1971). See also Robert Boguslaw, Tile New Utopians: A Study of System Design and Social Change (Englewood Cliffs, N. J. : Prentice-Hall, 1965); Joseph Bensman and Robert Lilienfeld, Craft and Con- sciousness: Occupational T echnique and the Development of W orld Images (New York: Wiley, 1973), pp. 282 ff; Robert Lilienfeld, "Systems Theory as an Ideology," Social Research 42 (Winter 1975): 637~60.
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES 145
is the difference itself: the difference of the present future and future presents. This difference itself is a historical fact, pro- duced and reproduced by the structure of our society. W e cannot avoid it or circumvent it as long as we continue to live in this highly complex society. But this does not mean that we have to pursue these pointless polemics.
Still, critical discussion and polemics have the important ad- vantage of being present behavior. Any attempt to replace them by posing the problem of temporal integration would defer the solution of this problem into the future and would, thereby, slide off into either utopian or technical channels. Again, the prob- lem of temporal integration, too, would become either a utopian or a technical problem and, thus, perpetuate itself.
An open and indeterminate future seems to suggest a shift from cognition to action, as Marx would have it, or today from pre- dicting to creating the future. 41 This sounds like: If you can- not see, you have to actl But both, prediction and action, have their utopian and their technical aspects. Substituting the one for the other does not solve the problem of temporal integration. The complex society of our day has to use both ways for reducing the complexity of its future; it has rather to sequentialize predic- tions and actions into complex self-referential patterns. There is no problem of choice between prediction and action, but there may be a problem of social and structural limitations for the com- bination of predictions and actions.
Social Communication as a Nontemporal Extension of Time
It should be clear by now that we can expect temporal integra- tion and, for that matter, integration of utopian schemes and tech- nology only as a present performance. Therefore, older societies which thought of themselves as living in an enduring or even
U So Bettina J. Huber, "Some Thoughts on Creating the Future," Sociological In- quiry 44 (1974): 29-39.
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
146 SOCIAL RESEARCH
eternal present did not experience our problem. Only in modern times. and only by shortening the time span of the present. does the problem of perseverance. or conservatio, get its actuality. 42 and only then do utopian schemes and technology diverge. By re-
structuring time in the last 200 years. the present has become specialized in the function of temporal integration; however. it
does not have enough time to do this job.
It is at this point that we can grasp the importance of the
theoretical contributions of George Herbert Mead 48 and Alfred Schutz44 concerning the interrelations between temporal and social experience. Both authors were aware of the fact that social communication defines the present lor the actors (because it com- mits the actors to the premise of simultaneity) and provides in addition the chance lor a nontemporal extension 01 time. "The field of mind. " in the words of Mead. "is the temporal extension of the environment of the organism. " and the mechanisms which accomplish this are social ones. 41i But then. the environment of systems can be also used as a nontemporal extension of time.
Other persons are socially relevant only insofar as they present. in communication. different pasts and/or different futures. They transform in a highly selective way distant temporal relevances into present social ones. And it is this selectivity that can be sub- mitted to social control-for example. by the twin mechanisms of trust and distrust. 46 This nontemporal extension of time by com-
munication constitutes time horizons for selective behavior-that is. a past that can never be reproduced because it is too complex and a future that cannot begin. And it is again this temporal com-
42 Cf. Hans Blumenberg. Selbsterhaltung und Beharrung: Zur Konstitution der neu%eitlichen Rationalitiit (Abhandlungen der Akademie der Wissenschaften und der Literature in Mainz. Wiesbaden, 1970).
48 Mead, The Philosophy 0 / the Present.
44 See above all Alfred Schutz, Der sinnha/te Au/bau der so%ialen Welt (Vienna: J. Springer, 1982).
411 Mead, The Philosophy 0/ the Present, p. 25.
40 For a more extensive treatment, see Niklas Luhmann, P'ertrauen: Ein Mech-
anismus der Reduktion so%ialer Komplexiliit, 2nd ed. (Stuttgart: F. Enke, 1978).
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES
147
plexity that makes selectivity necessary for meaningful behavior and communication.
These considerations bring us back to the roots of evolutionary interdependencies between social and temporal structures. Since this can be regarded as achieved knowledge, we cannot afford to
fall back on much simpler notions of the future as most social fore- casting does. The conception of interdependency, however, is in itself too vague and indeterminate to serve as a framework for further analysis. Neither Mead nor Schutz had adequate suc- cessors. The next step, indeed, is a difficult one. It requires the conceptualization of limitations and of gains that might result from novel combinations.
In view of the facts our society has produced in its bourgeois phase we should be able to calculate the limits of the meaningful extension of time; we should know the social correlates of a high differentiation of temporal horizons; we should be able to antici- pate a change in temporal structures as a consequence of social change-for example, as a consequence of an eventual decline of the monetary mechanism; we should be able to estimate the degree of heterogeneity of temporal structures we can tolerate in different subsystems of our society; we should know how the shrinking temporal horizons of families affect the economy, and how we can avoid the well known negative impact which the time perspectives of a growing economy have on the political system; 47 and, last but
not least, we should know what is implied if we rely on clocks and dates to integrate the different time perspectives of different sectors of the society and what dysfunctional consequences we have to expect if we use chronology to fulfill this important function.
It is sure that we cannot reduce this set of complex questions, involving the future, to a single one: how to begin the future. It is difficult to see how we could proceed in elaborating these questions or even answering them. Systems theory seems to be the only conceptual framework which has sufficient complexity.
61 For a classical statement, see Alexis de Tocqueville, L'Ancien regime et la revolu- tion. 5th ed. (Paris, 1866).
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
?
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p.
130
148 SOCIAL RESEARCH
So far, however, systems theory has used only very simple, chrono- logical notions of time and future, conceiving of the future simply as the state of the system at a later time. 48 Only environment, but not time, is recognized as a set of possible restraints on system states. Abstracting from time is, of course, quite legitimate as a scientific procedure; but then we must refrain from using tempo- ral notions in presenting the results.
In comparison with the conceptual elaboration of problems of time, systems theory 'is much more advanced in its conceptual com- plexity. It is the theory of time that is lagging behind, not the theory of systems. Not only social science but also the theory of history suffers from this deficiency. If the theory of time could be advanced, there Inight appear highly suggestive possibilities of research in correlations between system structures and temporal
structures.
The theory of time has to transform its vague idea of "every-
thing is possible in the long run," based on a chronological con- ception of time, into a concept of temporal structures with limited possibilities of change. It is a prerequisite of correlations that both variables are reduced contingencies in the sense that they cannot assume any shape whatever. We have, therefore, to look for time-inherent restrictions of possible correlations (substituting this for older notions of the substance or essence of time) before we set out to establish correlations between system structures and temporal structures. These time-inherent restrictions are, never- theless, results of sociocultural evolution and not a priori as- sumptions about the nature of the world or conditions of cogni- tion.
If we conceive of time as the relation between (more or less differentiated) temporal horizons and if we use a conceptual lan- guage that allows for iterative modalizations (present future, fu-
ture presents, future of past presents, etc. ) and define the function of the present and the function of chronology in these terms, we
48 See as a rather typical example Ervin Laszlo, A Strategy for tile Future: The Systems Approach to World Order (New York: Braziller, 1974).
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES 149
may have a sufficient base to start this kind of research. But we have to remain aware of the fact that a commitment to these con- ceptualizations is a commitment to "modern times. " Older so- cieties did not produce such an elaborated framework, and they did not need it to understand themselves. They lived, for struc- tural reasons we may be able to explain, within a less differentiated time.
The Future of Systems
Social systems are nontemporal extensions of tilne. They make the time horizons of other actors available within one contempo- rary present. This requires for social systems a double relation to time: a sequential one conceivable as process or as action in terms of means and ends, and a structural one conceivable as the differ- ence between system and environment. With respect to time, the difference of system and environment means that no complex sys- tem can rely exclusively on point-to-point relations to its environ- ment-that is, on instantaneous adjustment by immediate experi- ence and immediate reaction. 49 It needs time for its own opera- tions. This presupposes that under normal conditions no single event will change the whole system at once. Changing everything at once amounts to destruction. In other words: There is no con- ceivable state of a complex system which could be achieved by changing everything at once. The structural technique by which a system avoids this condition of changing everything at once is differentiation-or more exactly: a matching of internal and ex- ternal differentiation. lSo It is only at this rather taxing theoretical level of the relation between the relations of system/environment
49 Cf. Talcott Parsons, "Some Problems of General Theory in Sociology:' in John C. McKinney and Edward A. Tiryakian, eds. , Theoretical Sociology: Perspectives and Developments (New York: Appleton? Century? Crofts, 1970), pp. 27-68.
110 Cf. W. Ross Ashby, Design for a Drain (New York: WHey, 1952). Cf. also Uriel G. Foa, Terence R. Mitchell, and Fred E. Ficdlcr, "Differentiation Matching," Be- hav;oral Science 16 (1971): 130-142.
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
150 SOCIAL RESEARCH
and structure/process that we are able to locate our problem. Systems, then, in relation to their environment, depend for tem- poral reasons on a differentiation of structure and process. The
time perspective of modern society, on the other hand, project,s the difference of the present future and future presents. Both distinctions, worked out in very different intellectual traditions, seem to converge. If this is true, we can bring together systems
theory and phenomenological research.
In fact, the process of continuing communication in social sys-
tems under the condition of contemporaneity is the prospect of sequential social presents that will constitute forever new futures and new pasts. They are and will remain presents because they require a simultaneous integration of the perspectives of different actors. Structure, on the other hand, establishes for our society an open future in the sense that it provides for the selectivity of future presents.
Stated in more concrete terms, structure makes it possible and even necessary to postpone choices and to use the present future as a kind of storehouse for decisions to be made later. At the same time, the present system operates on the premise of continuing its processes. As a system it reproduces its present step by step. This sequentializing of presents, however, is meaningful only as a chain of choices, not as a chain of facts. The process of communication has its effect in producing and reproducing choice situations.
Going further, we have to break up this general notion of post- ponement of choices and have to distinguish two essentially differ- ent forms: (I) deferment of gratification and (2) deferment of ne-
gation. Both have their functional and institutional correlates. Deferment of gratification is a main prerequisite for the economic system as a condition for capital investment. Deferment of nega- tion is a main prerequisite of the political system as a condition of trust in political power. Both require institutional support, both require a present future for their present motivation. Both re- quire a working integration of utopian schemes and technology
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
TEMPORAL STRUCTURES 151
and a kind of security base for trust. III Both would not survive a considerable shrinkage of time horizons. Both may be endan- gered by too high a fluctuation rate of utopian schemes and tech- nological innovations. And, last but not least, do we not take too much for granted that it is and will remain possible, in spite of changing structural conditions, to separate deferment of gratifica- tion and deferment of negation and to avoid spill-over effects? Or will a refusal to defer gratifications any longer amount to a re-
fusal to defer negations; and finally, will the shrinking of time horizons in the economy endanger trust in politics, political ideol- ogies, value schemes, etc?
All of these questions pertain to what we have come to call bilrgerliche Gesellschaft and relate to the continuity or discon- tinuity of its structures under changing conditions. The bilrger- liche Gesellschaft has been a revolutionary society with a strong structural emphasis on time and corresponding simplifications of social and environmental relations. The principle of its future was simply the denial of its past 112 by the antistructural postulate of equality. lls The self-conception of this society"in its bourgeois variant did rely heavily on time-using and time-binding mecha- nisms like money and legal procedure. By now, we are aware of
111 For the function of security bases in relation to generalized media. of communi- cation. see Talcott Parsons, "On the Concept of Power" ~md "On the Concept of InOuence. " in his Sociological Theory and Modern Society (New York: Free Press, 1967). pp. 297-354. 355-382. Furthermore, Niklas Luhmann. "Symbiotische Mech- anismen," in Otthein Rammstedt, ed. , Gewaltverhiiltnisse und die Ohnmacht der Kritik (Frankfurt: Suhrkamp, 1974), pp. 107-131.
112 Cf. Joachim Ritter, Hegel und die franzosische Revolution (Koln: W est? deutscher V erlag, 1957).
113 A well known statement is Antoine de Condorcet, Esquisse d'un tableau his- torique des progres de l'esprit humain (1794). For the continuing impact of this idea and for empirical correlations between future orientation and emphasis on equality. see James A. Mau. Social Change and Images of the Future: A Study of the Pursuit of Progress in Jamaica (Cambridge: Schenkman. 1968). Since equality implies freedom and freedom implies inequality. the postulate of equality cannot refer to reality. but only to time. Its only function is to deny the relevance of the past--e. g. , the relevance of biographies and ascribed status for the access to educa-
tion (equality of opportunity) or to political elections.
Copyright (c) 2000 Bell & Howell Information and Learning Company Copyright (c) New School of Social Research
? Luhmann, Niklas, The Future Cannot Begin: Temporal Structures in Modern Society , Social Research, 43:1 (1976:Spring) p. 130
152 SOCIAL RESEARCH
highly complex operating conditions and of the narrow limits of effectiveness these mechanisms are subject to. In its Marxist or dialectical variant, the theory of society has to build its concept of future on negations of the present; but there is much more to negate in our present society than dialecticians could ever use for constructing or even bringing about one and only one de- sirable future: They have to focus on one central problem, thus overstating centralization, and to discount complexity in order to design a strictly linear theory which can be used to reconstruct or even to change the "process of history. "
There are many reasons, then, to suspect that the burgerliche Gesellschaft went very far in temporalizing reality and that the twin conceptions of bourgeois and Marxist theory were based on this. common presupposition. This does not decide the question whether this is a temporary distortion characteristic of the period of transition into a new type of world society, or whether this reflects lasting prerequisites of highly complex societies and/or an acceleration of the evolutionary process without parallels in pre-
vious history. We are certainly not prepared to decide this ques- tion without further research on the conceptual as well as on the empirical level. But we have the intellectual resources to go be- yond the boring controversies of Marxist versus- bourgeois or
utopian versus technocratic theory, and the starting positions are available for working out a systems theory of society which recog- nizes the fact that the future cannot begin and which compensates by the higher complexity of its conception of time for what might appear as a loss of future.
-I am indebted to S. Ho)mes. S. Seldman. and A. J. Vidich for comments on an earlier draft of this paper.
