The talk presented results from:
1. "Consistency Conditions for Orientifolds and D-Manifolds," E. G. Gimon and J. Polchinski, hep-th/9601038
2. "Anomalies, Dualities, and Topology of D=6 N=1 Superstring Vacua," M. Berkooz, R. G. Leigh, J. Polchinski, J. H. Schwarz, N. Seiberg, E. Witten, hep-th/9605184
3. "Tensors From K3 Orientifolds," J. Polchinski, hep-th/9606165
The main points were:
1) The world-sheet consistency conditions for orientifolds, as developed in refs. [1] and [3]. (Transparency 6 below contains some additional references.)
2) While world-sheet consistency guarantees perturbative consistency, there are now many examples of string theories which are perturbatively consistent but non-perturbatively inconsistent [2]. The point is that in some gauge backgrounds tensor representations (say of SO(32)) are well-defined but spinor representations are not (they do not satisfy the Dirac quantization on some 2-sphere, for example). Open string perturbation theory in such a background will be entirely non-pathological because the fundamental strings are all tensors. Nonperturbatively, however, the theory has D-branes carrying spinor representations and so is inconsistent.
Scanned Transparencies: 1| 2| 3| 4| 5| 6| 7| 8
Back to Strings '96 main page. Back to my home page.
FAQ:
How do you know the D-branes are in the spectrum?
The simplest and best answer is the consistency of the intricate web of ideas in string duality. Without a nonperturbative formulation of the theory one cannot prove anything, but if such an audacious set of conjectures were incorrect it would quickly become evident. Instead, they survive determined efforts to disprove them.
A more direct and purely perturbative argument can also be made. Start from the Type I string. By a series of T-dualities and continuous deformations (such as turning on gauge fields), one can obtain an arbitrary configuration of D-branes. The usual Type I string is an expansion around a state with 32 9-branes. The usual IIA and IIB strings are expansions around states with no branes. By purely perturbative reasoning one concludes that all these theories are different backgrounds in what is usually call the Type I theory, as are the same theories plus any configuration of D-branes. It could be that there is another Type II theory without D-branes, but Occam's Razor would suggest that a given perturbative string theory is not likely to have two completely different nonperturbative completions.
So does a skeptic also have to believe in string duality?
Yes. At strong coupling, the BPS/Witten-Olive argument implies that D-branes in the Type I, IIA, and IIB string theories are far lighter than the Planck scale. This implies in turn a large set of light string or Kaluza-Klein states from the dual theory. In the absence of a nonperturbative formulation one can believe anything, but at this point one has to lean over backwards to disbelieve string duality.
Updated July 26, 1996