Last spring there was a conference held in Florence which brought together many of those who worked on dual models and string theory during the late sixties and early seventies. Slides from the talks are here, and many of the speakers have written up contributions that have been posted on the arXiv. The latest of these is From Dual Models to String Theory, by Peter Goddard, who now is the director of the IAS in Princeton. It contains a detailed description of what he remembers of those early days when people were trying to sort out the significance of the Veneziano amplitudes and how to consistently quantize the string. Goddard also has some interesting remarks on the rapid changes in fashion during those years, and some excerpts follow.
On his student days at Cambridge working under Polkinghorne during the late sixties:
I, and nearly all my fellow research students, worked on strong interaction physics. (One of us was trying to work out the correct Feynman rules for gauge field theories, but this tended to be regarded as a rather recondite or eccentric enterprise.)
At a summer conference in 1971:
For me, it was a memorable meeting and one particular vignette has stuck in my mind as an illustration of the prevailing attitude towards the use of modern mathematics in theoretical high energy physics. A senior and warmly admired physicist gave some lectures on the Regge theory of high energy processes. With great technical mastery, he was covering the board with special functions, doing manipulations that I knew from my studies with Alan White (who was also at the School) could be handled efficiently and elegantly using harmonic analysis on noncompact groups. Just as I was wondering whether it might be too impertinent to make a remark to this effect, the lecturer turned to the audience and said, “They tell me that you can do this all more easily if you use group theory, but I tell you that, if you are strong, you do not need group theory.”
About his years (1970-72) at CERN:
The two years I had spent in CERN had built up to an crescendo of intellectual excitement and, though I have found much of my subsequent research work gripping and often extremely satisfying (when teaching duties and the largely self-inflicted wounds of administration have permitted), nothing has quite matched this period. In particular, I had the privilege of working closely for seven or eight months with Charles Thorn, whose combination of deep perception and formidable calculational power had provided the basis of what we managed to do. And, the exhilarating combination of the open and cooperative atmosphere that prevailed amongst (almost all) those working on dual models in CERN, the relative youth of most of those involved, the sense of elucidating a theory that was radically different, even the frisson of excitement that came from doing something that was regarded by some of those in power as wicked, because it might have nothing directly to do with the real world – this cocktail would never be offered to me again.
About the situation in 1973-4 , after the discovery of asymptotic freedom:
By the end of 1973, as the fascination of dual models or string theory remained undimmed, though with ever increasing technical demands, the interest of many was shifting elsewhere. On 21 December, David Olive wrote to me, “Very few people are now interested in dual theories here in CERN. Amati and Fubini independently made statements to the effect that dual theory is now the most exciting theory that they have seen but that it is too difficult for them to work with. The main excitement [is] the renormalization group and asymptotic freedom, which are indeed interesting.”
In Berkeley [summer 1974], I wrote a largely cathartic paper  on supersymmetry, which probably helped no one’s understanding, except marginally my own. It had one memorable effect: namely, that when I reached Princeton I was invited to give a general seminar on supersymmetry, which most people did not know much about then. When I said I would rather talk about string theory, my offer was politely declined on the grounds that no one in Princeton was
interested, a situation that has changed in the intervening years. Somewhat put out by this response, I did not give a seminar at all.
About his decision in 1975-6 to work on gauge theory rather than strings:
I started to realize that following my interests in strings or dual models might be a fine indulgence for me, but it was not going to help my students get jobs. (One of the great attractions of Cambridge at the time was that chances for promotion were so slim – Jeffrey Goldstone was still a Lecturer – that one did not need to be distracted by the prospects for advancement: they seemed negligible.)