We list these conferences and will report on those that had the most productive outcomes, with results that are still central to our understanding of information structures in the universe and biological information processing.

1. 1946, March

   Core Group: Bateson, Bigelow, von Bonin, Frank, Fremont-Smith, Gerard, Harrower, Hutchinson, Klüver, Kubie, Lazarsfeld, Lewin, Lorente de Nó, McCulloch (chair), Mead, von Neumann, Northrop, Pitts, Rosenblueth, Savage, Wiener
   Guest: Bremer

   
   

   • Inaugural Macy Conference entitled "Feedback Mechanisms and Circular Causal Systems in Biological and Social Systems."

     The conference title will shift repeatedly during the series. In addition, Wiener's invocation of the term 'cybernetics' for the subject matter was still in the future.

     
     
   • Macy Foundation organizer Frank Fremont-Smith opens the conference (and the ensuing series) by saying, "Each group, when it comes together, is an experiment. If it excites you all enough to want to meet again, we will plan for further meetings."

     It is important to bear in mind that to Fremont-Smith the conferences themselves were an experiment in multidisciplinary science.

     
     
   • Because the 1st conference represented the first 'coming together' of the hard scientists and the social scientists, the discussion was more general and more philosophical than would be the case in the subsequent 9 conferences.
      
     
   • The conference's opening session had von Neumann and Lorente de Nó presenting detailed overviews of the state of the art in digital computers and neurophysiology, respectively.
      
     
   • The conference's second session had Wiener presenting an overview of automatic mechanisms for self-regulation. Rosenblueth then described purposive behavior and teleological mechanisms a la his 1942 presentation.
      
     
   • McCulloch gives a presentation on how simulated neural networks can emulate the calculus of propositional logic. He also draws attention to communication as a descriptive metaphor and notes the differences between descriptions of messages' mechanics and message content or meaning. He suggests memory may be a function of continuously cyclical impulses in a neural network.
      
     
   • Bateson makes a presentation outlining the need for sound theory in the social sciences, illustrating his points with observations from his anthropological field work of the 1930's. He distinguishes between 'learning' and 'learning to learn', then challenges the group by asking whether and how computers could accomplish either form of learning.
      
     
   • Wiener and von Neumann in particular make claims that their theories and models would be of utility in economics and political science.

     No scholar-representative from economics or political science would ever attend any of the 10 conferences.

     
     
   • Gerard comments that the brain's operations are much more 'analog' than 'digital'.

     This establishes a dichotomy between 'analogical' and 'digital' which would become a recurrent topic of debate throughout the conferences. Some (especially the mathematicians like von Neumann) would be emphasizing 'digital' perspectives, while others (especially the psychologists) would be emphasizing a more 'analogical' orientation.

     
     
   • Heinrich Klüver gives a presentation on how object perception appears to use feedback mechanisms to enforce constancy. He declares psychology lacks a good model explaining how a brain handles form perception (Gestalten) - posing a topic which would be addressed repeatedly in the next few conferences.
      
     
   • Lawrence Frank suggests the conference's focal interdisciplinary concepts could only be elaborated and developed with the aid of a language more general than the disciplinary lexicons of the day.

     This allusion to the need for a 'meta-language' would be repeated again and again throughout the conference series. Fremont-Smith would open each conference with a reminder that it was necessary to establish a new lexicon or language for the new ideas being discussed.

     
     
   • Psychologist Molly Harrower gives a presentation on perceptual differences between normal persons and people with brain damage.
      
     
   • Psychiatrist Lawrence Kubie gives a presentation on neurosis, emphasizing compulsive repetitive behaviors in neurosis. Kubie's descriptive allusions to 'energy' (a la Freud) set off much discussion.

     Such (often critical and animated) discussion on psychiatric issues and models would recur throughout the conference series. Kubie in particular would demonstrate remarkable persistence and patience in making regular presentations on psychoanalytical topics which would invite criticism from the 'hard scientists' like Pitts.

     
     
   • Northrop - the sole philosopher in the core group - gives a presentation on philosophy of science. He brings up the notion of an ethics derivable from science and recommends that there be attention given to generating a valid normative theory grounded in scientific principles and evidence. He discerns a lack of interest in these philosophical issues among his fellow group members, and rarely speaks out after this first conference.

     Northrop's appearance illustrates three things. First, not all participants' contributions were treated with equal interest by the group at large. Second, not all core group members actively contributed presentations throughout the conference series. Third, although some would treat the Macy Conferences as 'philosophical' and as an 'exercise in metascience', Northrop's experience would seem to indicate neither of these themes struck a chord with the audience at large.

     
     
   • Sociologist Lazarsfeld proposes that a separate meeting be arranged specifically for social scientists, so as to provide a forum for introducing (e.g.) social theorists to these new concepts.

     This proposal (which was acted on) reinforces some scholars' opinion that the essence of the Macy Conferences was to impart the new ideas to a primarily social science audience.

     
     
   • The group agrees to meet again in a second such conference to occur in October.

     The participants found the inaugural conference sufficiently enlightening and interesting to motivate a commitment to attend further such conferences.

     
     

   ("Summary of First Macy Conference,"American Society for Cybernetics.)

2. 1946, October

   Core Group: Bateson, Bigelow, von Bonin, Frank, Fremont-Smith, Gerard, Harrower, Hutchinson, Klüver, Kubie, Lazarsfeld, Lewin, Lorente de Nó, McCulloch (chair), Mead, von Neumann, Northrop, Pitts, Rosenblueth, Savage, Wiener, + Brosin, + Marquis, + Schneirla
   Guest: Livingston
   

   • Conference now entitled "Teleological Mechanisms and Circular Causal Systems."
      
     
   • As a result of Lazarsfeld's suggestion at the 1st conference, the Macy Foundation has Gregory Bateson organized a 'special conference' (Dupuy) or 'sociological subconference' (Heims) held in September. The title of this meeting was 'Teleological Mechanisms in Society'. It was designed to allow social scientists to meet with Wiener and von Neumann, to hear about their ideas, and to discuss how these ideas might be valuable in social science.

     This 'side event' turned out to be noteworthy on its own. Two sociologists who would later invoke systems principles - Talcott Parsons and Robert Merton - attend this subconference (but never attend any of the main Macy Conferences). This event was the first Macy appearance by anthropologist Clyde Kluckhohn (who would be a guest at 2 of the main conferences). This subconference recommended to the group at large that the concepts of 'field' and 'Gestalt' be clarified, and this recommendation is acted upon.

     
     
   • The conference takes up the clarification of the terms 'field' and 'Gestalt'.

     The main outcome of this discussion is illustration of how little the attendees agreed on the definitions and implications of these labels. They end up deferring further discussion until the seminal Gestalt psychologist Wolfgang Köhler can address the conference. (NOTE: Köhler was invited to the 3rd conference, but couldn't actually attend until the 4th).

     
     
   • Molly Harrower is asked to explain the term 'field', but she demurs by noting her mentor Koffka's ideas diverged from those of Köhler.
      
     
   • Kurt Lewin gives an extensive presentation on his personal version of 'field' and other concepts from Gestalt psychology and social psychology.
      
     
   • Three psychologists are added to the core group.

     These first additions to the core group increase the already-disproportionate representation of psychologists among the conference attendees.

     
     
   • Schneirla (a comparative psychologist - i.e., an 'ethologist') gives a presentation on tactile and chemical communications within an army ant society.

     This is one of the rare occasions where psychological or behavioral issues were contextualized in terms of a collective (as contrasted with individuals).

     
     
   • Lawrence Frank arranges a separate conference on 'teleological mechanisms' conducted under the auspices of the New York Academy of Sciences and held in New York immediately following the 2nd Macy Conference. From the Macy attendees Frank draws the featured speakers Wiener, McCulloch, Hutchinson, and Livingston.

     The 2nd conference should therefore be seen as the central event in a trio of meetings. The earlier 'sociological subconference' and this latter conference serve to expand the audience for the Macy Conferences' core themes outside the context of the main conference itself.

     
     
   • In the wake of the 2nd conference, both Bateson and Northrop contact chairperson McCulloch to express reservations about the wisdom of focusing so much on Köhler. Bateson was concerned about the multidisciplinary conference getting diverted into intra-disciplinary controversies. Northrop saw Gestalt psychology as a suboptimal theme, and recommended keeping the focus on 'harder' science such as (e.g.) neurophysiology and mathematics.

     This illustrates the extent to which (a) there was occasional individual trepidations about the topical and procedural foci of the conference series as well as (b) there were a lot of back-channel communications concerning the conduct of the conferences.

   ("Summary of Second Macy Conference,"American Society for Cybernetics.)
3. 1947, March

   Core Group: Bateson, Bigelow, von Bonin, Frank, Fremont-Smith, Gerard, Harrower, Hutchinson, Klüver, Kubie, Lazarsfeld, - Lewin, Lorente de Nó, McCulloch (chair), Mead, von Neumann, Northrop, Pitts, Rosenblueth, Savage, Wiener, + Brosin, + Marquis, + Schneirla
   Guests: Ackerman, Erikson, Festinger, Fitch, Kluckhohn, Lloyd
   

   • Conference still entitled "Teleological Mechanisms and Circular Causal Systems."
      
     
   • For the first time, the number of invited guests is allowed to exceed the 5 originally planned by Fremont-Smith.
      
     
   • McCulloch begins collating and distributing a summary report on the conference

     This illustrates the realization that the earlier conferences had not been adequately documented. McCulloch attempted to summarize the key points of the first 3 conferences following the third event, then distributed this to the attendees. With the exception of Mead's shorthand notes (indecipherable owing to her personal shorthand coding) and the fragmentary results of the earlier crude recording / transcription efforts, McCulloch's retrospective remains the main documentation for the first 3 conferences.

     
     
   • First loss of core group member - Kurt Lewin dies shortly before the conference
      
     
   • Lewin's death cancels planned appearance by Köhler to discuss his view of perception (distinct from the 'coding' model previously discussed)
      
     
   • Erik Erikson gives a presentation on child psychiatry. His approach is considered non-rigorous compared with the tone of the conferences to date. Although Bateson and Hutchinson lobby for Erikson's inclusion in the core group, opposition prevents this. For his own part, Erikson was uncomfortable with the group's focus on machines.

     This illustrates the problems (both internal and external) in adding new core members to the group. It also illustrates the fact that not every new attendee saw the conference and/or its subject matter as something attractive enough to pursue.

     
     
   • In relation to his own research interests (e.g., cellular automata), von Neumann lobbies to get a geneticist invited to the conferences.

     This is the first important instance in which von Neumann acts not only on his own initiative, but with regard to his own personal topical interests. In retrospect, von Neumann would later be characterized (e.g., by von Foerster) as something of an 'insider' (member of the core group) who operated more like an 'outsider' (i.e., someone pursuing his own agenda tangential to that of the collective).

   ("Summary of Third Macy Conference,"American Society for Cybernetics.)
4. 1947, October

   Core Group: Bateson, Bigelow, von Bonin, Frank, Fremont-Smith, Gerard, Harrower, Hutchinson, - Klüver, Kubie, Lazarsfeld, - Lewin, Lorente de Nó, McCulloch (chair), Mead, von Neumann, Northrop, Pitts, Rosenblueth, Savage, Wiener, + Brosin, + Marquis, + Schneirla

   Guests: Teuber, Bender, Kluckhohn, Garcia Ramos
   

   • Title of the conference is modified to "Circular Causal and Feedback Mechanisms in Biological and Social Systems."

     This represents the third title in four events. It illustrates the manner in which the conference participants were still at a loss to provide a specific denotation for what they found so interesting.

     
     
   • Köhler presents his 'field' perspective, generating substantial controversy. Pitts and McCulloch criticize Köhler's 'field theory' as mere theory devoid of empirical basis.

     This debate illustrates two important points. First, there was a divide between those interested in the mechanisms of neural architectures (e.g., Pitts and McCulloch) and those who expressed interest in descriptive theories of what those mechanisms might do (e.g., perception). Second, it remained the case throughout the conference series that the neural mechanism people (who were also to be counted among the formalists) were repeatedly critical of what they perceived as fuzzy or vapid theorization (particularly with regard to the psychoanalysts).

     
     
   • Long argumentation over the distinctions between the continuous or 'analog' character of Köhler's Gestalt model and the discretely-coded or 'digital' orientation adopted by (e.g.) McCulloch and Pitts.
      
     
   • Harrower and Teuber are disappointed at poor reception given Köhler's presentation. Harrower threatens to resign from the core group, but is persuaded to stay.

     This illustrates that the Macy Conference audience was, after all, a collection of people, each with his or her own personal commitments and attachments. The social network of these commitments and attachments would influence the conferences in many ways - including selection of new invitees and the departure of some people along the way.

   ("Summary of Fourth Macy Conference,"American Society for Cybernetics.)
5. 1948, Spring

   Core Group: Bateson, Bigelow, von Bonin, Frank, Fremont-Smith, Gerard, Harrower, Hutchinson, - Klüver, Kubie, Lazarsfeld, - Lewin, Lorente de Nó, McCulloch (chair), Mead, von Neumann, Northrop, Pitts, Rosenblueth, Savage, Wiener, + Brosin, + Marquis, + Schneirla, + Teuber, + Bavelas
   Guests: Jakobson, Lee , Lotz, Morris, von Domarus, Delbrück (geneticist selected by von Neumann, expected to become member of core group)
   

   • Title of the conference remains "Circular Causal and Feedback Mechanisms in Biological and Social Systems."
      
     
   • First day's program, organized by Mead and Bateson, focuses on language

     Although this is not the only conference in which 'language' is a topic of presentation and discussion, it is the only conference in which an entire block of presentations on this topic was presented. It is unclear whether or not this attention to language overshadowed the invited presentation by biophysicist / geneticist Delbrück - a type of scholar and a specific individual recommended by von Neumann.

     
     
   • Second day's program dominated by presentations by Wiener (order vs. chaos), Pitts (formal modeling applied to chicken pecking order formation), and Lee (concept of "I" in language).
      
     
   • Geneticist Delbrück unimpressed by fifth conference - later stating, "It was vacuous in the extreme and positively inane." This illustrates the difficulty in getting new attendees to integrate themselves into the conference and/or the group.

     Delbrück was not the first or the only invitee to think little of his Macy Conference experience (cf. the earlier comments about Erikson). His reaction illustrates that not everyone readily accepted or endorsed the conferences (though the precise reasons remain unclear). It must be noted that the dedication of a significant block of this conference's itinerary to 'language' may have diminished the attention given Delbrück and the genetics subject matter which von Neumann had been promoting, and that this might help explain his negative reaction. It must also be pointed out that Delbrück's statement was given to Steven Heims in the 1960's, and that time may have amplified his displeasure.

     As was the case with Erikson, Delbrück had been invited with some expectation of his being added to the core group. As was the case with Erikson, Delbrück found himself sufficiently uncomfortable with the conference as to prevent his further attendance, much less his incorporation into the 'inner circle'. This illustrates that recognition of the conferences' themes was not universal.

   ("Summary of Fifth Macy Conference,"American Society for Cybernetics.)
6. 1949, March

   Core Group: Bateson, Bigelow, von Bonin, Frank, Fremont-Smith, Gerard, - Harrower, Hutchinson, - Klüver, Kubie, - Lazarsfeld, - Lewin, Lorente de Nó, McCulloch (chair), Mead, - von Neumann, Northrop, Pitts, Rosenblueth, Savage, Wiener, + Brosin, + Marquis, + Schneirla, + Teuber, + Bavelas, + von Foerster (guest - invited to join)
   Guests: Abramson, Liddell, Lindsle,y Lloyd, Mettler, Stroud
   

   • The conference opens with discussion of a message from (non-attendee) von Neumann. Having calculated the number of neurons and interneuronal connections in the brain he'd claimed the brain's neurons were insufficient to account for human capacities, and that the means for achieving the brain's 'complexity' must include other mechanisms such as the biochemical structure of the neuron itself. The physiologists present were pleased at this idea. McCulloch defended the viability of his and Pitts' neuron model (which had thus been called into question). The debate ends when Pitts demonstrates von Neumann's statements had been based on a calculation which was invalid.

     This event is interesting for a number of reasons. First, although von Neumann operated as something of an 'outsider' and wasn't even present, the attendees still took the time to consider his message. Secondly, this could have turned into another in the recurring rounds of debate over neural mechanisms versus subjective perception had not Pitts short-circuited the discussion by pointing out the flaw in von Neumann's description of the problem.

     
     
   • Klüver suggests research topic of analyzing situations leading to childhood trauma.
      
     
   • During this period, McCulloch becomes a severe and vociferous critic of psychoanalysis - alienating some of the psychological and psychiatric acquaintances who participate in the Macy conferences.
      
     
   • Kubie makes note of the (problematical) role of the observer in psychoanalytical work, and claims the therapist has to remain as detached as possible, even to the exclusion of humanistic impulses. Wiener brings up the problems of measurement interfering with observed phenomena in the sciences. Fremont-Smith and Stroud join in.

     This is the only occasion when the problem of the observer is explicitly discussed in the Macy Conference series. This seems strange in retrospect, given that the problem of the observer would become so prominent two decades later. It is perhaps less strange when one considers the fact that discussion of 'subjective experience' (cf. earlier debates on Gestalten) tended to be promoted only by the 'soft science' crowd (particularly the psychoanalysts), and that they were repeatedly criticized by the core 'hard science' advocates for doing so. In any case, it's fair to say this resistance to the topic of subjective experience would force deferral of this critical topic to the 1960's and the rise of second-order cybernetics.

     
     
   • Heinz von Foerster's presentation on memory politely received. He's invited to become editor for the Conference proceedings (allegedly so he can practice English). Mead and Teuber are appointed as assistant editors. These assignments persist for the remainder of the conference series.
      
     
   • Admittedly daunted by the length and complexity of the conference title, von Foerster recommends Wiener's recently-published label 'cybernetics' be adopted as the conference title. This is enthusiastically approved. Wiener, deeply touched, leaves room to hide his tears.

     This anecdote (cited in multiple of von Foerster's documented reminiscences) illustrates how concise circumscription of the conference's subject matter under a discrete label had not theretofore been given much attention. More generally, it's indicative of the inattention given to portraying the conference's themes as a discernible field in and of itself.

     
     
   • Fremont-Smith makes an appeal for collaboration between physics and psychology (and by implication all the 'hard' and 'soft' sciences) leading to unification of science. He emphasizes the need for cross-disciplinary awareness and declares, "The development of effective communication across the scientific disciplines in perhaps the most urgent need of our era."

     This illustrates multiple points. First, Fremont-Smith (as mentioned earlier) was very interested in these conferences serving as the birthplace for a truly multidisciplinary (or transdisciplinary) field (or metascience). Second, as of halfway through the eventual series, there was little evidence this was occurring. Third, it illustrates that Fremont-Smith was growing increasingly concerned at the extent to which the attendees continued to engage each other in terms of their respective disciplinary perspectives.

   ("Summary of Sixth Macy Conference,"American Society for Cybernetics.)
7. 1950, March

   Core Group: Bateson, Bigelow, von Bonin, Frank, Fremont-Smith, Gerard, - Harrower, Hutchinson, - Klüver, Kubie, - Lazarsfeld, - Lewin, Lorente de Nó, McCulloch (chair), Mead, - von Neumann, Northrop, Pitts, Rosenblueth, Savage, Wiener, + Brosin, + Marquis, + Schneirla, + Teuber, + Bavelas, + von Foerster (guest - invited to join)
   Guests: Licklider, McLardy, Shannon, Stroud, Werner
   

   • In accordance with von Foerster's suggestion (cf. 6th conference notes), the conference title becomes "Cybernetics: Circular Causal and Feedback Mechanisms in Biological and Social Systems."

     It is important to note that the Macy Conferences were not titled or characterized as 'cybernetics' until 6 of the 10 conferences had already occurred. At this late stage, the conference title is stabilized once and for all.

     
     
   • Atypically large proportion of presentations by guests (instead of core group members).
      
     
   • Gerard starts conference with presentation on 'analog' versus 'digital' interpretations of mind. He states the mind is more toward the 'analog', calling into question the 'digital' logic-based model of Pitts and McCulloch. This sets off an animated debate that proves frustrating to many of the participants.
      
     
   • Bateson calls for clarification of distinction between 'analog' and 'digital'. He hearkens back to the arguments over Köhler's presentation at the 4th conference and suggested it would be wise to remove any ambiguities.

     It is interesting to note that the old debate concerning 'analogical versus digital' remained a pesky item of 'old business unresolved'. Gerard's points are essentially the same that he'd made years earlier. Bateson's allusions to Köhler's presentation only remind us that the prior attempt to invoke and understand Köhler led only to dissension and hurt feelings and not any progress toward understanding with respect to Gerard's distinction.

     
     
   • Remainder of the conference is largely dedicated to presentations on communication and language.

     At the 5th conference there was a block of time invested in presentations on human language. This time around, the theme wasn't so much language in and of itself, but rather how human language intersected with the features relevant to (e.g.) Shannon's information theory. In this regard, it's important to note that this 7th conference was the first of three that Shannon himself would attend.

     
     
   • Psychoanalyst Kubie gives a presentation on language and symbols as they relate to neurosis. This touches off a debate with Pitts and Bateson criticizing psychoanalysis from two distinct perspectives. Pitts can't discern any coherent theory in psychoanalysis, while Bateson can't discern any objectivity in psychoanalytic tenets.

     This illustrates two points. The first was that no amount of criticism from the 'hard science' contingent (especially Pitts) would dissuade Kubie from presenting psychoanalytical issues at the conferences. The second was that some people from the 'soft science' contingent (in this case Bateson) also had problems accepting psychoanalytic theory or terminology as sufficiently clear or rigorous.

     
     
   • Guest Licklider gives a presentation on analyzing 'intelligibility' in speech communications - a topic quite distinct from reducing uncertainty a la Shannon's information theory. This sets off a series of exchanges wandering through topics like (e.g.) emotional tone and the speech of parrots.
      
     
   • Claude Shannon presents a paper on a formal analysis of semantic redundancy in printed English, focusing on writing as coding and tilting the discussion from the social scientists (energized by Licklider) over to the engineers.
      
     
   • Gerard makes critical remarks about overinflated claims for cybernetics and undue publicity given recent conferences. He muses that the group had originally presented ideas freely and uncritically, but had become apparently overconfident about their theories. He goes on to warn about the risk of accepting and relying upon mathematical models incapable of empirical validation through observation and experimentation.

     Gerard had served as a recurrent gadfly during the conferences (cf. his repeated observations on the 'analogical versus digital' distinction). This marked the first time anyone had pointed to the manenr in which these conferences' subject matter and participants were being externally viewed and treated as a cohesive group with a coherent agenda. His remarks were motivated in large part by articles in the popular press (e.g., Time) that were characterizing this 'cybernetics group' as the vanguard of an imminent fusion of man and machine. His remarks about mathematical modeling and empiricism are also interesting, because they mark the first recognition of the risks inherent in relying only on abstract models as the entirety of their methodology.

     
     
   • Psychologist Klüver criticizes both Köhler's 'field theory' and McCulloch's 'digital' models with respect to perception. He suggests both are too abstract to constructively analyze (e.g.) the functionality of the visual system in the course of 'seeing'.

     This illustrates how little progress had been made in response to Klüver's original challenge to the 1st conference to explain 'Gestalten'. Although the group had gone out of their way to invite Köhler so as to resolve such issues, that had ended up being a disruption rather than an illumination. It also highlights the fact that in between the engineering of artifical neural models and the fuzzy subject of psychoanalysis was a middle ground of experimental and comparative psychology which had been underrepresented and little noted during the conferences to date.

   ("Summary of Seventh Macy Conference,"American Society for Cybernetics.)
8. 1951, March

   Core Group: - Bateson, Bigelow, von Bonin, Frank, Fremont-Smith, Gerard, - Harrower, Hutchinson, - Klüver, Kubie, - Lazarsfeld, - Lewin, Lorente de Nó, McCulloch (chair), Mead, - von Neumann, Northrop, Pitts, Rosenblueth, Savage, - Wiener, + Brosin, + Marquis, + Schneirla, + Teuber, + Bavelas, + von Foerster (guest - invited to join)
   Guests: Birch, Bowman, MacKay, Rioch, Richards, Shannon
   

   • Neither Wiener nor von Neumann attended this 8th or any subsequent Macy Conference.

     These two 'dropouts' are significant in light of the fact these individuals would later be commonly cited as instrumental to and central in the Macy Conferences and in cybernetics generally. It is fair to say that their roles and criticality has been overstated in later decades (particularly in the case of von Neumann, the 'outsider').

     
     
   • MacKay's views on information (distinct from Shannon's by virtue of trying to incorporate 'meaning') are evident in his presentation, which touches off a debate on whether meaning is an intrinsic component of 'information' as that construct is being used at the time.

     One can see this debate over 'meaning' and 'information' as analogous to the earlier debates over 'subjective experience' versus 'neural mechanisms'. In both cases, some participants wanted to address things solely in terms of cognitive 'vehicle', while others wanted to address cognitive 'content.' By and large, the 'vehicle-only' adherents prevailed throughout the conference series.

     
     
   • MacKay also suggested automatons could be capable of inductive inference if configured to employ random strategies. This brings a severe criticism from statistician and decision theorist Savage, who claims randomness adds nothing in emulating human behavior and can only diminish problem solving efficiency.

     This debate is interesting in the sense that what MacKay was proposing can be seen as the basis for exploration, adaptation, and learning. Given the prominence that learning and education would achieve in later decades (e.g., the work of Gordon Pask), it's ironic that 'learning in systems' would be such a rare subject of discussion in the Macy Conferences.

     
     
   • Savage gives a presentation on his seminal decision theory research, based on statistical analysis and predicated on a quantifiable 'utility' metric. McCulloch responds critically, arguing that decision contexts aren't typically reducible to any such one-dimensional metric.
      
     
   • Bavelas presents some of his recent experiments in small group dynamics and group communications.

     Bavelas' presentation is interesting because it illustrates how the Macy Conferences influenced the thinking and methods among some of the 'soft science' participants. This case illustrates Bavelas' drift toward concise and abstract theoretical elements (a la information theory) and away from the positions he'd held when he first appeared at the 5th conference. When he arrived, he could be characterized as an adherent of Lewin and Köhler's Gestalt approach. By this time, he'd begun to jettison some of the more 'fuzzy' descriptive concepts and make an effort to adopt more rigorous constructs.

     
     
   • Discussion of Bavelas' structured (and game-like) experiments runs the gamut from applicability of von Neumann's game theory to psychic motivations to ESP to applicability of 'meaning' within an 'information theory' to anxiety about automatic machines to the role of machines as models in understanding human cognition.

     This illustrates that even at this late date the conferences continued to engender stimulating and multi-faceted conversations.

     
     
   • Literary critic Richards gives a presentation on the type of language necessary to address and analyze language itself.

     The inconclusive discussion that followed Richards' presentation includes the only known allusion to Gödel's incompleteness theorem in the entire Macy series. Given the mathematical expertise present in the conferences and the subsequent prominence of Gödel's theorem in critiques of artificial intelligence, this is something of a surprise.

     
     
   • Animal communications researcher Herbert Birch makes a presentation in which he draws a distinction between interactions that are merely 'behavior' versus what he terms 'true communication' in higher animals and humans. This 'true communication' Birch characterizes as involving anticipation, intentionality, symbolization, learning, and social engagement. Rosenblueth and Bigelow respond very critically, claiming that the elements of Birch's 'true communication' entail ambiguous mentalist notions that are inimical to the non-mentalist approach underlying their paradigmatic cybernetics research.
      
     
   • In an exchange on the recurring topic of the 'unconscious', Rosenblueth declares that a mental event (and / or associated neural events) either occurs or it does not occur. He characterized the notion of a mental event engendering a memory but somehow still 'unconscious' to be nonsense.
      
     
   • As the recurring debate over psychiatry's status as 'scientific' played out, Rosenblueth claims natural science's general approach and language can handle the problems addressed by psychiatry, and Pitts claims that the onus is on psychiatrists to either demonstrate their methods are 'scientific' or else advance them to where they become so.

     These three factoids provide evidence of something that had permeated the entire conference series - a persistent critique from the 'hard science' contingent against fuzzy descriptions based on vague 'mentalist' constructs. This in turn illustrates why the participants interested in cognitive 'content' (cf. Klüver's persistent allusions to Gestalten) were always at a disadvantage in putting such topics before the group.

   ("Summary of Eighth Macy Conference,"American Society for Cybernetics.)
9. 1952, March

   Core Group: - Bateson, Bigelow, von Bonin, Frank, Fremont-Smith, Gerard, - Harrower, Hutchinson, - Klüver, Kubie, - Lazarsfeld, - Lewin, Lorente de Nó, McCulloch (chair), Mead, - von Neumann, - Northrop, Pitts, Rosenblueth, - Savage, - Wiener, + Brosin, + Marquis, + Schneirla, + Teuber, + Bavelas, + von Foerster (guest - invited to join)
   Guests: Ashby, Bowman, Luce, Monnier, Quastler , Remond, Torre , Wiesner, Young
   

   • More 'guests' invited than at any other Macy Conference.
      
     
   • McCulloch opens the conference with observations about increasing disruptions to the conference series' coherence. He cites the problems of conflicting schedules which had forced a number of regulars to be elsewhere. He also cites the growing secrecy imposed on some participants' projects (von Neumann and Bavelas) which prevents presentation and discussion of relevant work.

     By this time von Neumann had effectively dropped out of the conference series. McCulloch's remarks illustrate both the increasingly oppressive atmosphere of the McCarthy era and the apparent fact that long-term conference participants were prioritizing other things.

     
     
   • Divergent interests continue to characterize the participants' discussions. For example, Bateson responds to Gerard's presentation on neuronal excitation / inhibition by asking how one can relate such neurophysiological minutiae to broader philosophical or epistemological issues.
      
     
   • Bateson gives a presentation on humor and communication, leading to his unveiling of a notion that paradox (his espoused key to humor) was at the heart of all human communication.
      
     
   • W. Ross Ashby presents 2 papers - one on his 'homeostat' and the other on the prospects of chess playing automatons requiring random tactics before they can defeat human opponents.

     Though he would later become one of the most famous cyberneticians, this was Ashby's only appearance in the Macy Conference series.

     
     
   • Interestingly, some of the more technically-minded regulars like Bigelow and Pitts, interrogate Ashby about his homeostat and challenge him to explain how it 'learns'.

     Ironically, this means the 'hardcore' cyberneticians were in the position of arguing their case from an internalist or mentalist perspective - the same position for which they'd been castigating the psychiatrists for years. It's also ironic in the sense that such critiques had for years stifled attention to the subject of 'learning' - the very point upon which they challenged Ashby.

     
     
   • Kubie avoids some of the arguments associated with his earlier presentations by downplaying psychoanalytical theory and characterizing himself as a sort of naturalist observing emotions.

     This illustrates a drift similar to Bavelas (cf. 8th conference notes), but one more like a defensive reaction rather than a positive adaptation based on past discussions.

     
     
   • Guest Quastler proposes application of cybernetics at the microlevel in relation to biochemical and cellular processes. He presents a set of estimated values for 'complexity' in biological organisms, based on the amount of information they represent or can represent.

     Quastler's presentation apparently didn't generate a lot of discussion or enthusiasm. Interestingly, the notion of quantifying 'complexity' would serve as the seed for 'complexity studies' - one of the many alleged latter-day successors to cybernetics.

     
     
   • Bigelow chides the social science contingent for falling into a trap by surrendering to a fascination with mathematics.

     Dupuy (2000) cites this as a specific illustration of cybernetics' problem with managing non-specialists' and public perceptions of the usually very specific concepts and constructs the group generally addressed. In this case, Bigelow was warning the non-specialist participants that they were uncritically accepting some of the 'hard science' contingent's products without understanding the limits of their significance. This also relates to Gerard's earlier comments warning of overinflated expectations among the public.

   ("Summary of Ninth Macy Conference,"American Society for Cybernetics.)
10. 1953, April

    Core Group: Bateson, Bigelow, von Bonin, Frank, Fremont-Smith, Gerard, - Harrower, Hutchinson, - Klüver, Kubie, - Lazarsfeld, - Lewin, Lorente de Nó, McCulloch (chair), Mead, - von Neumann, - Northrop, Pitts, Rosenblueth, - Savage, - Wiener, + Brosin, + Marquis, + Schneirla, + Teuber, + Bavelas, + von Foerster (guest - invited to join)
    Guests: Amassian, Bar-Hillel, Bowman, Chao, Droogleever-Fortuyn, Quastler, Shannon, Grey-Walter
    

    • This was the only Macy Conference held somewhere other than the Beekman Hotel in New York City.

      Ironically, this final conference was supposedly moved to Princeton for the convenience of von Neumann, who for all intents and purposes had dropped out of the 'cybernetics group' three conferences earlier.

      
      
    • Discussion at this conference is unusually animated in style and ambiguous in content. Assistant editor Teuber described the conference as lacking content, and threatens to resign if anyone pushes to publish a proceedings for such an affair. In a compromise to prevent Teuber's resignation, it is agreed that only papers (and not discussion transcripts) will be included in the final transactions, which will not appear until 2 years later.

      This illustrates how the long-term conference attendees were showing signs of fatigue.

      
      
    • McCulloch reports on his and Pitts' work on how neural mechanisms can recognize shapes and musical chords. He cites strong arguments from others rebutting this work, and ends with a good-natured concession that his and Pitts' efforts have been in the fine tradition of scientific refutability.

      This incident is a poignant event, in that at the time it seemed to indicate one of the project streams feeding the Macy Conferences had in fact turned out to be a dead end.

      
      
    • McCulloch is tasked to write a final summarization of the consensus achieved during the 10 Macy Conferences. This proves difficult, because by this time it's clear that the cybernetics group is moving (and has always moved) in several different directions. McCulloch writes in part: "Our most notable agreement is that we have learned to know one another a bit better, and to fight fair in our shirt sleeves." (Transactions, p. 69)

      As chairperson for all 10 Macy Conferences, McCulloch no doubt desired to portray the series as having produced something. His concession of what can only be called a social networking outcome illustrates how the Macy Conferences could not even then be construed as having produced a unified theory or meta-discipline of the sort to which Fremont-Smith had aspired.

    ("Summary of Tenth Macy Conference,"American Society for Cybernetics.)

1. 1950, September

   At the first London Symposium on Information Theory, held at the Royal Society, about twenty papers were presented, including six on applications of information theory to psychology and neurophysiology.

2. 1952, September

   At the second symposium, held at the Institution of Electrical Engineers, emphasis on psychology and neurophysiology was replaced by an emphasis on the transmission and analysis of speech, with which eight of the thirty-eight papers dealt.

3. 1955, September

   Participants (selected): J. T. Allanson, Yehoshua Bar-Hillel, D. E. Broadbent, E. C. Cherry, Peter Elias, Dennis Gabor, R. L. Gregory, Victor Ingve, J. Y. Lettvin, J. C. R. Licklider, Benoit Mandelbrot, W. S. McCulloch, Margaret Mead, W. H. Pitts, Oliver G. Selfridge, Henry Quastler, A. S. C. Ross, The scope of the third London Symposium on Information, held at the Royal Institution, was broadened to include not only all of these fields but the mechanical translation of languages as well, in addition to the basic topics of information theory, coding, etc.

   The breadth of its scope is also indicated by the range of backgrounds of the participants, which included: anatomy, animal welfare, anthropology, computers, economics, electronics, linguistics, mathematics, neuropsychiatry, neurophysiology, philosophy, phonetics, physics, political theory, psychology, and statistics. Of the 250 participants, half were British; one-eighth came from the United States (Shannon was not among them); the remainder, in order of decreasing numbers, were from the Netherlands, Sweden, France, Germany, Denmark, the USSR, Italy, Belgium, Switzerland, Spain, and Israel.

   Thursday’s session began with “The Place of ‘Meaning’ in the Theory of Information,” by Dr. D. M. Mackay of Kings College, London. Mackay defined the “meaning” of a message in terms of the change it produces in the recipient’s “conditional probability matrix” (array of probabilities of anticipated events). In response to a question about “aesthetic information” and the “meaning” of music, Mackay indicated he was willing to include the effects on the recipient’s internal secretions, etc. Bar-Hillel stated that he had a hard time getting anything out of Mackay’s talk and felt it was pointless to define “meaning” in terms of even vaguer matrices. A. S. C. Ross also questioned Mackay’s ideas.

   Professor J. C. R. Licklider of M.I.T. spoke on “Auditory Frequency Analysis.” He discussed the incompatability of the place and frequency theories of pitch, and. he discussed Huggins’s demonstration of the appearance of a pitch when white noise is fed to both ears, arriving at one ear through an all-pass filter; the pitch that is heard is related to a frequency characteristic of the filter, although neither ear alone detects any pitch. This phenomenon requires phase or time analysis and cannot be explained by strict place theory; i.e., by a theory which postulates that every pitch is associated with a particular “place” in the cochlea. By means of a tape recording, Licklider demonstrated that the low pitch associated with a set of three or so of its neighboring higher harmonics can be detected even in the presence of low-pitched thermal noise that completely masks an equally loud fundamental. He concludes that the ears perform running analyses of three types: (1) spectrum, (2) autocorrelogram, and (3) cross-correlogram, and that through a process of association of nerve paths, a person comes to associate the agreeing reports of the three analyses concerning pitch. As a result, one has the sensation of a pitch when the envelope of a higher-frequency complex wave has the periodicity associated with a sine wave that pitch. However, if the phase relations of the harmonics are altered, the sensation of a low pitch may disappear because of the change in the shape of the envelope. Experiments have eliminated the possibility that the low pitch results from nonlinearity and beats. Licklider’s theory appears to offer the first unified explanation of a number of auditory phenomena. Afterward, Schouten also discussed and demonstrated the phenomenon of the missing fundamental, using three sine waves between 2 and 3 kc, separated by 200 cps and 200 cps to produce the sensation of 200 cps.

   Thursday’s session ended with a talk by R. L. Gregory of the Cambridge University Department of Experimental Psychology on “An Experimental Treatment of Vision as an Information Source and Noisy Channel.” He found an empirical extension of Weber’s law ΔI/I = C, where I is the level of illumination and ΔI is the smallest distinguishable increment, to include the effect of the areas A₁, and A₂ of the contrasting fields, namely,

   ΔI/I = C₁ + C₁√(1/A₁ + 1/A₂)

   He regards the breakdown of Weber’s law at low intensities as due to “noise” and finds that the law may be extended to cover low intensities by rewriting it in the form ΔI/(I + k) = C. The constant k lies between 0.03 and 0.04 foot-lamberts, but its relationship to the “noise” is not clear. Gregory also attempted to find a function describing the observed relation between area and threshold intensity on the basis of a threshold defined in terms of the probability of mistaking noise for a signal, etc. In the discussion period, Licklider pointed out that Tanner at the University of Michigan asserts there is no sensory threshold-merely a point below which the probability of detection is low. Licklider reported that if an image is stabilized on the retina, visual acuity is momentarily improved, but the image then disappears on account of fatigue. Small movements of the eye therefore do not appear to be responsible for the acuity.

   On Friday, J. T. Allanson of the University of the Birmingham University Electrical Engineering Department began the last day’s session with a talk on “Some Properties of a Randomly Connected Neural Network.” In order to permit a mathematical discussion of the behavior of such a network, he postulated a much simplified model, and derived expressions for the equilibrium rate of firing of neurons, the rate of damping of oscillations, etc., in a fixed randomly connected network, attempting to relate the results to various neural phenomena. Because many rather different networks exhibit very similar gross behavior, Allanson concluded that electroencephalography is not likely to yield detailed information about the structure of the central nervous system.

   P. D. Wall delivered a paper by himself, J. Y. Lettvin, W. S. McCulloch, and W. H. Pitts, all of M.I.T., on “Factors Limiting the Maximum Impulse-Transmitting Ability of an Afferent System of Nerve Fibers.” Experiments were conducted: 1) to determine the maximum sustained frequency of nerve impulses which can be carried along a set of fibers originating in the skin, muscles, and leg joints, proceeding into the spinal column, and streaming toward the head in the dorsal columns of the spinal cord; 2) to determine the limits on impulse transmission following a single impulse or short burst of impulses where all the impulses are carried by the same set of nerve fibers; and 3) to determine the limits following activity in a neighboring parallel set of nerve fibers. The object of the investigation is eventually to calculate the informational capacity of such a system and to ascertain how close to it the rate of transmission of information in the nervous system comes. If it is close, information theory should provide insights for neurophysiology. The quantity of greatest interest is the mean frequency of impulses required to achieve the channel capacity, which is to be compared with the actual average frequency of impulses.

   After lunch, Dr. Oliver G. Selfridge of the Lincoln Laboratory, M.I.T., presented a very interesting paper “Pattern Recognition and Learning.” The type of learning to which he referred is the evolution of a method for distinguishing patterns, which first requires a motivation and some sort of criterion (i.e., elementary method) for pattern recognition.

   Next D. E. Broadbent of the Applied Psychology Research Unit, Cambridge, England, under the title “The Concept of Capacity and the Theory of Behavior,” presented a plea for the qualitative use of information theory in psychology, “to provide a language for talking about events within the man’s nervous system.” In the discussion period, the question of “semantic hygiene” was raised in connection with this proposed use of exact terms in new and vague ways.

   After the final tea interval, Henry Quastler of the University of Illinois Control Systems Laboratory spoke on ‘Studies of Human Channel Capacity.” He has found in experiments with reading a single line of random piano music involving an alphabet of 3 to 65 keys in which all notes are equally long, a maximum information rate of 23 bits per second is achieved with a 25-key alphabet. (No key is allowed to follow immediately after itself.) Physical limitations reduce the information rate when the alphabet size exceeds about fifty keys or the required speed exceeds about 5.2 keys per second. Quastler found that the addition of more lines, accompaniment, words, or varied rhythm to the music did not permit a higher information rate. This is to be contrasted with Licklider’s discovery of a total rate of 35 bits per second for combined reading and pointing which separately can achieve rates of 25 and 15 bits per second, respectively.

   A five-minute talk in English by V. Siforov, one of the Russians present, expressed his appreciation for the symposium and the bringing together of people from different countries and sciences; he also expressed his appreciation for Shannon’s work. It was his feeling that information theory will yield new high-speed means of communication. He reported that among those in the Soviet Union working in the field of information theory were: Kolmogorov; Khinchin; Yagvo (?); Patelnikov, who works on the theory of potential stability; Kharkevich, who is working on the accommodation of signals and the extension of Shannon’s work (he is reported to have published recently in Radiotechnika); Sardoniko, who works on automatic regulation; and the speaker himself, who works on the detection of weak pulsed signals and on active noisy networks. Finally, he expressed his appreciation for the invitation to the symposium sent to his country and for the attention of the audience.

   ("The Third London Symposium on Information Theory*,"IRE TRANSACTIONS-INFORMATION THEORY , March, 1956.)

• 1952

  One of the basic tools in natural science is the energy concept. In recent years, another concept has begun to attain comparable dignity. It is something more subtle and elusive than energy; it is derived from a desire for dealing methodically with problems of complexity, order, organization, specificity.... It is known as entropy or amount of information, and plays a prominent role in the new fields of information theory, communication theory, and cybernetics.

  The "new movement" is based on evaluative concepts (R. A. Fisher, s experimental design, A. Wald's statistical decision function, J. von Neumann's theory of games), on the development of a measure of information (R. Hartley, D. Gabor, N. Wiener, C. Shannon), on studies of control mechanisms, and the analysis and design of large systems (W. S. McCulloch and W. Pitt's "neurons," J. von Neumann's theory of complicated automata, N. Wiener's cybernetics).

  It has become increasingly evident that the principles of information theory are applicable to the "higher" functions of living organisms. They can also be used to advantage in analyzing basic functions such as metabolism, growth, and differentiation. There are remarks to that effect in Wiener's Cybernetics; in E. Schroedinger's What is Life? there is a provocative discussion on how organisms feed on "negative entropy." In the present volume, the formalism of information theory and cybernetics has been applied to some basic biological problems, with some degree of success. The amount of control and communication involved in the basic, biological processes is very great. Indeed, it appears that there is little difference between the exercise of the basic biological activities, and the exercise of the most sophisticated skills in the most highly differentiated organisms. We do not claim that the absolute difference between basic and highly specialized activities is small; after all, this difference is the result of the order of a billion years of organic evolution. The claim is only that the gap between simple and complicated living organisms is small in comparison to the gap which separates the simplest living things from the most complicated non-living systems, say, between a bacterium and a giant electronic brain. Thus, when we say that the functional complexities of man and bacterium are closely comparable, we do so, not because of a low estimate of the functional complexity of man, but because we are very much impressed by the complexity of bacterial life.

  The cybernetics of biological functions should ultimately contribute to the understanding of the nature of biological control systems. The first stage in the analysis is quantification: a count of control elements, a measure of the amount of control exercised. Most papers in this volume deal with counts and measures; only on a few occasions have steps beyond this stage been attempted.

  It is in the nature of cybernetics that it cuts across traditional boundaries between various fields of science. Because of this, symposia have played a major part in its development--in particular, the British symposia ('50, '52), the series of Macy conferences, the French symposium ('52). The present volume, too, is the result of a cooperative effort; it stems largely from lectures and discussions arranged, in the summer of 1952, under the auspices of the Control Systems Laboratory at the University of Illinois.

  (Information Theory in Biology, 1953, Henry Quastler, ed.)

• 1956

  Information theory is based on the concept that information is measurable. This idea is not new. In physics, the notion of a measurable relation between information and degree of orderliness (entropy) dates back to BOLTZMANN'S work in 1872 and its development in 1929 by SZILARD (1). In 1918, the statistician R. A. FISHER (2) needed a criterion to assess the degree to which the information contained in experimental data is utilized by a given statistical procedure; he worked out a measure of information which has been used in statistics ever since. Later, the need arose for a measure of information carrying potential as a consequence of the tremendous development of telecommunication, and in 1928, R. V. L. HARTLEY (3) published such a measure.

  In 1948, WIENER (4) observed that a measure of information content is a basic ingredient to the study of communication, which itself is a basic ingredient to the study of control in its broadest sense. In the same year, the communication engineer C. E. SHANNON (5) published an article on the mathematical theory of communication which in several respects went beyond previous studies. This article is highly technical; it is very difficult reading; it appeared in a specialized journal (The Bell Systems Technical Journal) and it pertained to no other field than telecommunication. It certainly did not look like an article destined to reach wide popularity among psychologists, linguists, mathematicians, biologists, economists, estheticists, historians, physicists . . . yet this is what happened. In 1949, the University of Illinois Press issued a book (6) which consisted of a reprint of SHANNON'S earlier article and a paper by WARREN WEAVER ; in this paper, the generality of the concept of 'amount of information' was forcefully expounded. The literature on 'information' has been increasing ever since at an almost explosive rate. Gatlinburg, Tennessee

  (Symposium on Information Theory in Biology, 1958, Hubert Yockey, Ed.)

• 1966-67 International Union of Biological Sciences Symposia - "Towards a Theoretical Biology"

  Theoretical physics is a well recognized discipline, and there are departments and professorships devoted to the subject in many universities. Moreover, it is widely accepted that theories of the nature of the physical universe have profound consequences for problems of general philosophy. In contrast to this situation, theoretical biology can hardly be said to exist as an academic discipline. There is even little agreement as to what topics it should deal with, or in what manner it should proceed.

  At the beginning of this century it was usually considered that the most characteristic feature of life is its metabolism. "The constant synthesis, then, of specific material from simple compounds of a non-specific character is the chief feature by which living matter differs from non-living matter", was the way it was expressed by Jacques Loeb in 1916. At that time, there was a good deal of discussion about whether living systems presented us with real or only apparent and local exceptions to the Second Law of Thermodynamics.

  About 40 years ago, however, it began to be realized that such views gave insufficient attention to the overwhelmingly important phenomenon of evolution and the increase in complexity of living things. The view gradually gained ground, particularly under the urging of H. J. Muller, that the essential feature of life resides in its possession of a hereditary system capable of mutation, and thus of being affected by the process of natural selection. The orthodox view became, and in many quarters perhaps still remains, that the basic elements of life are the genes as units of information.

  The older view should not be completely abandoned. One can find examples of mutable hereditary information in extremely simple systems that no one would consider alive. Two examples were fairly thoroughly discussed; irregularities in the crystal lattices of complex materials such as clays by Cairns Smith, and tactic copolymers by Pattee

  No conceptualization of a living system is adequate unless it includes at least four importantly different time scales, those of metabolism, development, heredity and evolution. In those parts of the discussion more concerned with biology than the deeper philosophical questions of quantum theory and the like, all these four levels came in for some consideration. The processes of metabolism were considered mainly in their more fundamental aspects, that is to say, in connexion with the synthesis of new compounds, particularly proteins. The discussions started from the two basic points; that even the simplest living systems are exceedingly complex, so that many synthetic processes are proceeding simultaneously; and that each of these processes is subject to control mechanisms, often of the type loosely referred to as negative feedback. One of the major methods used in the physical sciences for the handling of complex systems is statistical mechanics.

  At the second symposium in particular, there was an extended discussion of the applications of statistical mechanics to a variety of biological problems, not all at the level of metabolism.

  At the second meeting Richard Gregory discussed a theory of perception under the provocative title "How so little information controls so much behaviour". His answer was, roughly, that a small amount of information arriving through the sense organs activates "pre-existing" models in the brain—which can be otherwise expressed by saying that the incoming, information falls within the domain of a certain chreod and thus converges on to its attractor.

  Evolution was perhaps the most central theme throughout the whole discussions. Many physicists seem ready to concede that the principle of natural selection imparts to the biological world a type of logical structure, which they scarcely meet in their own field of interest. Biologists, however, while gratified to be told that physicists admit that biology offers problems which actually need thinking about, still remain doubtful whether physicists have realized just how challenging these problems really are... historical detail, For reasons which it would be invidious to go into in any historical detail, Neo-Darwinism has become an established orthodoxy, any criticism of which is regarded as little less than lese-majéste.

  The discussion [brought] one face to face with problems which offer as much intellectual challenge as quantum indeterminacy or Bohr's complementarity. Theorists in general science have staked out claims for a variety of fields in this area—games theory, decision theory, systems theory, and the like. It is not clear to the biologists wrestling with actual situations that many of these "disciplines" amount to anything more than the formulation of a lot of problems for which no solutions can be provided; there seems to be a singular dearth of actually proved theorems which the biologists can take over and employ. Possibly the people who are trying to discover how to set up a computer to learn to play good chess, or bridge, are among those most likely to make a major contribution to the fundamental theory of evolution.

  ("Towards A Theoretical Biology," C. H. Waddington, Nature, V. 218. May 11. 1968 p.525, C. H. Waddington)

• 1974 - "Process Thought and Modern Science"

• 1956

  Dartmouth Summer Research Project

  In 1955, a summer study project at Dartmouth College was proposed by John McCarthy (with Marvin Minsky and Claude Shannon, who is said to have proposed it). The proposal is credited with introducing the term 'artificial intelligence'.

  “We propose that a 2 month, 10 man study of artificial intelligence be carried out during the summer of 1956 at Dartmouth College in Hanover, New Hampshire. The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it. An attempt will be made to find how to make machines use language, form abstractions and concepts, solve kinds of problems now reserved for humans, and improve themselves. We think that a significant advance can be made in one or more of these problems if a carefully selected group of scientists work on it together for a summer.”

  Attendees at the Dartmouth summer project included: W. Ross Ashby, Alex Bernstein, Julian Bigelow, Tom Etter, John Holland, Donald Mackay, John McCarthy, W.S. McCulloch, Marvin Minsky, Trenchard More, John Nash, Allen Newell, Abraham Robinson, Nathanial Rochester, Arthur Samuel, David Sayre, Oliver Selfridge, Claude Shannon, Kenneth R. Shoulders, Herbert Simon, and Ray Solomonoff, who gave us the most complete record of the workshop proceedings.

  The most complete record of the workshop was that of Ray Solomonoff, who, with founders John McCarthy and Marvin Minsky, were the only ones to attend the entire eight weeks.

  The absence at the Dartmouth Summer Project of the pioneer of cybernetics, MIT's Norbert Wiener, was most conspicuous. The conference proposal deliberately did not mention cybernetics. That was the basis of the famous Macy Conferences, which ran from the late 1940's to around 1960, and included several of the AI pioneers mentioned above. McCarthy wanted to avoid discussions of simple automata theory. By avoiding cybernetics, which focused on analog feedback, it meant he also avoided debates with the powerful Wiener.

  In the first two weeks of the summer workshop, Newell and Simon put forward the idea of a two-stage "General Problem Solver," which was partial inspiration for Daniel Dennett's "Valerian Model" of free will in 1978.