math0202194/SQS.tex
1: %&LaTeX
2: 
3: %%%%%%%%%% espcrc2.tex %%%%%%%%%%
4: %
5: % $Id: espcrc2.tex 1.2 2000/07/24 09:12:51 spepping Exp spepping $
6: %
7: \documentclass[twoside]{article}%amsart}%
8: \usepackage{fleqn,espcrc2}
9: 
10: \usepackage{amsfonts}
11: 
12: % change this to the following line for use with LaTeX2.09
13: % \documentstyle[twoside,fleqn,espcrc2]{article}
14: 
15: % if you want to include PostScript figures
16: %\usepackage{graphicx}
17: % if you have landscape tables
18: %\usepackage[figuresright]{rotating}
19: 
20: % put your own definitions here:
21: % ===========================================
22: %
23: %                New Symbols
24: %
25: % ===========================================
26: \newcommand {\supplus}{\mathop{{\supset}\llap{\raise 
27: 0.5pt\hbox{\normalfont\small+}\hskip 0.5pt}}} 
28: %semidirect sum, the ideal on the left
29: 
30: \newcommand {\subplus}{\mathop{{\subset}\llap{\raise 
31: 0.5pt\hbox{\normalfont\small+}\hskip 0.5pt}}}  
32: %semidirect sum, the ideal on the right
33: 
34: %
35: % ===========================================
36: 
37: \newcommand {\Cee}    {{\mathbb  C}}
38: \newcommand {\Fee}    {{\mathbb  F}}
39: \newcommand {\Kee}    {{\mathbb  K}}
40: \newcommand {\Nee}    {{\mathbb  N}}
41: \newcommand {\Oee}    {{\mathbb  O}}
42: \newcommand {\Pee}    {{\mathbb  P}}
43: \newcommand {\Ree}    {{\mathbb  R}}
44: \newcommand {\Zee}    {{\mathbb  Z}}
45: 
46: \newcommand {\bcdot}   {\mathbin{\hbox{\raise.4ex\hbox{\bf.}}}} % bold \cdot
47: \newcommand{\rmname}[1]
48:   {\expandafter\newcommand \csname #1\endcsname {{\rm{#1}}}}
49: 
50: \rmname{id}
51: \rmname{Hom}
52: \rmname{Mat}
53: \rmname{qet}
54: \rmname{st}
55: \rmname{tr}
56: \rmname{Span}
57: \rmname{str}
58: \rmname{Vol}
59: \rmname{vol}
60: 
61: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
62: 
63: \newcommand{\ttbs}{\char'134}
64: %\newcommand{\AmS}{{\protect\the\textfont2
65: %  A\kern-.1667em\lower.5ex\hbox{M}\kern-.125emS}}
66: 
67: % add words to TeX's hyphenation exception list
68: \hyphenation{author another created financial paper re-commend-ed
69: Post-Script}
70: 
71: % declarations for front matter
72: \title{On unconventional integrations and cross ratio on
73: supermanifolds}
74: 
75: \author{Dimitry Leites\address[SU]{Department of Mathematics,
76: University of Stockholm, \\
77: Kr\"aftriket hus 6, SU-106 91, Stockholm,
78: Sweden; mleites@matematik.su.se}%
79: \thanks{I am thankful to NFR and TBSS for financial support, to
80: J.~Lukiersky for hospitality, to B.~Zupnik, O.~Hudaverdyan and
81: A.~Vaintrob for inspiring discussions, V.~Serganova and P.~Grozman for
82: help.}}
83: 
84: \begin{document}
85: 
86: \begin{abstract}
87: The conventional integration theory on supermanifolds had been
88: constructed so as to possess (an analog of) Stokes' formula.  In it,
89: the exterior differential $d$ is vital and the integrand is a section
90: of a fiber bundle of finite rank.  Other, not so popular, but,
91: nevertheless, known integrations are analogs of Berezin integral
92: associated with infinite dimensional fibers.  Here I offer other
93: unconventional integrations that appear thanks to existence of several
94: versions of traces and determinants and do not allow Stokes formula. 
95: Such unconventional integrations have no counterpart on manifolds
96: except in characteristic $p$.
97: 
98: Another type of invariants considered are analogs of the cross
99: ratio for ``classical superspaces''.  
100: 
101: As a digression, homological fields corresponding to simple Lie
102: algebras and superalgebras are described.\vspace{1pc}
103: \end{abstract}
104: 
105: % typeset front matter (including abstract)
106: \maketitle
107: 
108: 
109: For the basics on Linear Algebra in Superspaces and Supermanifold
110: theory see \cite{D}; for notations and useful facts see \cite{Sch},
111: \cite{Ser}.  At the talk I also considered related issues partly
112: collected in \cite{GLS1}.  As compared with the talk, \S\S2, 3 are
113: new; they are a part of the talk given 10 years earlier \cite{LSV} but
114: yet unpublished.  Encouraged by Manin's selected examples \cite{Mn}
115: and recent results in classification of simple Lie superalgebras
116: \cite{LSh1} I decided to draw attention to these issues.
117: 
118: 
119: 
120: \section{INTEGRATION}
121: 
122: 
123: \subsection{Integration with Stokes' formula} In mid 1970's
124: J.~Bernstein and I discussed how to construct an analog of integration
125: theory on supermanifolds.  We had at our disposal (1) the differential
126: forms, i.e., functions polynomial in differentials of the coordinates,
127: the coefficients of these polynomials being usual functions and (2)
128: volume forms, the latter constituted a rank one module $\Vol$ over the
129: algebra of functions and under the change of coordinates the generator
130: $\vol(x(y))$ of $\Vol$ accrued the Berezinian (superdeterminant) of the
131: Jacobi matrix as the factor.  Each of the above notions had to be
132: carefully reconsidered in super setting because even the most
133: innocent-looking notions and theorems (e.g., the Foubini theorem)
134: displayed, in supersetting, funny signs at unexpected places, see
135: \cite{L}, v.  31.
136: 
137: On manifolds, one {\it can} integrate differential forms; on
138: supermanifolds, one can {\it not}: their transformation rule yields no
139: analog of determinant, except in the absence of odd parameters.  On
140: the other hand, one, clearly, can integrate elements of $\Vol$,
141: provided they are with compact support, of course.  But we wanted to
142: have some analog of Stokes' formula, and, therefore, needed (1)
143: elements of ``degrees'' lesser than that of volume forms, and (2) the
144: notion of the supermanifold with boundary to overcome the puzzle
145: demonstrated by ``Rudakov's example''; for solution see \cite{BL1}.
146: 
147: To have integration theory, one needs not only what to
148: integrate (the integrand), but over what (cycle), and orientation. 
149: The latter two notions turned out to be more involved than we
150: originally thought; Shander clarified this in his development of
151: integration theory, see \cite{Sh} and the details in \cite{L}.
152: 
153: Actually, what we had had was sufficient to construct the integration
154: theory desired: by setting $\Sigma_{-i}=\Hom_{\tt F}(\Omega^i, \Vol)$,
155: where ${\tt F}$ is the superspace of functions, we obtain a complex
156: dual to the de Rham one with $\Sigma_{0}= \Vol$ as forms of the
157: highest degree.  We called the elements of $\Sigma_{\bcdot}$ {\it
158: integrable} forms (the ones one can integrate) and described how to
159: integrate such forms in \cite{BL1}.
160: 
161: \subsection{Veblen's problem and Rudakov} We wondered for a while if
162: there is another integration theory with Stokes' formula, and to
163: investigate the options, considered the following problem: {\sl
164: describe all differential operators acting in the spaces of tensor
165: fields and invariant with respect to any changes of variables}. 
166: Indeed, the exterior differential (instrumental in Stokes' formula)
167: is, evidently, an invariant and, as is proven in \cite{R}, this is the
168: only invariant unary differential operator between spaces of tensor
169: fields whose fibers are irreducible $\mathfrak{gl}(n)$-modules with
170: vacuum vector.  So, in order to scan the options on supermanifolds,
171: it was necessary to list all such operators.
172: 
173: This problem (to list all invariant differential operators) goes back
174: to O.~Veblen (see \cite{GLS1} for a review).  For unary operators on
175: manifolds it was solved by Rudakov \cite{R} as a part of another
176: problem (description of irreducible vacuum vector modules over simple
177: Lie algebras of formal or polynomial vector fields).
178: 
179: \subsection{Unconventional integrations} 
180: A.~Shwarts and his students \cite{AS} attempted to integrate densities
181: and objects depending on higher jets of the diffeomorphism but all
182: their examples boil down to either pseudodifferential (\cite{BL2})
183: or integral forms.  Having obtained an analog of Rudakov's result for
184: the general vectorial Lie {\it super}algebra \cite{BL3}, we can be
185: sure that there is only one integration theory on supermanifolds {\it
186: provided the integration involves tensors}
187: $$
188: \hbox{{\it with irreducible finite
189: dimensional fibers}.}\eqno{(*)}
190: $$
191: 
192: The result of \cite{BL3} do not preclude, however, unconventional
193: integrations.  For tensors other than $(*)$ constructions \`a la
194: Shwarts may lead to an integration theory (perhaps, \cite{H} leads to
195: it).  Such a theory does exist and in \cite{LKV} the calculations from
196: \cite{BL3} are used to consider infinite dimensional fibers snubbed at
197: in \cite{BL3} for no reason except tradition.  It turns out that in
198: the spaces of such tensors there act invariant operators similar to
199: Berezin integral.  Next, observe that having stated that it is
200: impossible to integrate differential forms on supermanifolds, we
201: almost immediately published a paper \cite{BL2} showing, nevertheless,
202: how to do it if one is very eager to.  More exactly, one has to
203: consider {\it pseudodifferential forms}, i.e., functions nonpolynomial
204: in differentials.  Of course, there are no such functions on
205: manifolds.  Certain types of pseudoforms lead to new invariants ---
206: semi-infinite cohomology of supermanifolds; quite criminally, no
207: examples are calculated yet.
208: 
209: Here I consider still another type of ``integrations''.
210: 
211: 
212: \subsection{Supertraces and superdeterminants} From the very beginning
213: I wondered what if we stop insisting on having an analog of the
214: Stokes' formula?  What remains of the integration then?  Only the
215: Jacobian, one can say.  Since it is easier to deal with Lie
216: algebras than with groups, let me list analogs of trace for Lie
217: superalgebras.  Then, if the Lie superalgebra $\mathfrak{g}$ can be
218: exponentiated to a Lie supergroup, we can consider the analog of the
219: determinate defined via the formula
220: $$
221: \det \exp X= e^{\tr (X)} \quad\hbox{  for any } \; X\in\mathfrak{
222: g}.\eqno{(1)}
223: $$
224: In other words, I mean: 
225: 
226: 1) Let us consider the Lie superalgebras $\mathfrak{g}$ with a trace
227: also denoted by $\tr$ (i.e., $\tr([x, y])=0$ for any
228: $x,y\in\mathfrak{g}$); then for the role of $\Vol$ we can take tensor
229: fields of type $\tr$, its infinitesimal transformations being the
230: Cartan prolongation (see \cite{GLS1}) of the pair $(\id,
231: \mathfrak{g})$, where $\id$ is the ``standard'' or ``identity''
232: representation of $\mathfrak{g}$.
233: 
234: The prime example is provided by the Poisson Lie superalgebra
235: $\mathfrak{g}=\mathfrak{po}(0|2n)$.  Indeed, there is a parametric family
236: (quantization) of Lie superalgebras $\mathfrak{g}_{t}$ which at $t=0$
237: coincides with $\mathfrak{po}(0|2n)$ and $\mathfrak{
238: g}_{t}\simeq\mathfrak{gl}(2^{n-1}|2^{n-1})$ for $t\neq 0$, see 
239: \cite{LSh2}.
240: 
241: 2) From various points of view it is clear that $\mathfrak{gl}(n)$ has
242: at least two superanalogs: the ``simple-minded'' one,
243: $\mathfrak{gl}(n|m)$, and the ``queer'' one, $\mathfrak{q}(n)$.  On
244: $\mathfrak{q}(n)$, the supertrace vanishes identically but there are
245: specially designed for it its particular, queer, trace and
246: determinant.  Regrettably, the queer trace is odd and, therefore, to
247: describe the corresponding representation, we need odd parameters,
248: which causes extra difficulties.
249: 
250: So, still another versions of integration theory, if exist, are
251: related with the queertrace and its ``quasiclassical limit'' as
252: $t\longrightarrow 0$: the restriction of the above quantization is a
253: parametric family $\mathfrak{g}_{t}$ which at $t=0$ coinsides with
254: $\mathfrak{po}(0|2n-1)$ and $\mathfrak{g}_{t}\simeq\mathfrak{
255: q}(2^{n-1})$ for $t\neq 0$. In 1) and 2) the identity representation 
256: of $\mathfrak{po}(0|m)$ is the adjoint one and the Berezin integral 
257: serves as $\tr$.
258: 
259: 3) The analog of trace on the general vectorial algebra $\mathfrak{
260: vect}(m|n)$ is the divergence.  I do not know how to generalize
261: formula $(1)$ with divergence serves as $\tr$, so let me mention two
262: other, more obvious, analogs of $\tr$: (a) in characteristic $p$ such
263: analog exists, e.g., for Lie algebras of contact vector fields (but
264: not only), see \cite{St}; another one is provided by (b)
265: ``superconformal'' algebras of divergence-free series $\mathfrak{
266: svect}(1|N)$and the exceptions related with $N=4$ and $N=5$ extended
267: Neveu-Schwarz algebras, see \cite{GLS2}; these traces are of the same
268: parity as $N$.
269: 
270: \section{Jordan superalgebras}
271: 
272: I consider here certain algebraic structures associated with certain
273: selected ``classical superdomains''.  So far, nobody knows yet (as far
274: as I know) even a ``right'' definition of this basic notion, to start
275: with: for finite dimensional manifolds all is clear, for
276: supermanifolds we just consider the most easy to handle simple Lie
277: supergroups and their Lie superalgebras whereas the elusive ``right''
278: definition requires, perhaps, semisimple or almost simple Lie
279: superalgebras.  So we take the road of least resistance:
280: 
281: Unless otherwise mentioned the ground field is $\Cee$, the classical
282: superdomains are considered as quotients of {\it simple} or close to
283: them ``classical'' Lie supergroups modulo {\it certain} maximal
284: parabolic subsupergroups; for the list see \cite{S2}.
285: 
286: This paper is an attempt to tackle the following questions: What are
287: the criteria for selecting the above-mentioned subgroups among other
288: maximal ones?  Why cosets modulo other parabolic subsupergroups are
289: seldom considered in Differential Geometry whereas only these
290: ``other'' cosets are the main topic of study in analytical mechanics
291: of nonholonomic dynamical systems and in supergravity (\cite{Ma},
292: \cite{GL})?
293: 
294: In these questions ``super'' is beside the point, so we can very well
295: begin with manifolds.  The classical domains are distinguished among
296: symmetric spaces by the fact that the Lie algebra of the symmetry
297: group of any classical domain (Hermitian symmetric space) $M$ is a
298: simple complex Lie algebra of the form 
299: $$
300: \mathfrak{g} = \oplus_{|i|\leq 1}\; \mathfrak{g}_i;\eqno{(d=1)}
301: $$ 
302: the tangent space to $M$ at a fixed point can be identified with the
303: $\mathfrak{g}_{-1}$ and, on it, one can always define a Jordan algebra
304: structure, by fixing any element $p\in\mathfrak{g}_{1}$ and setting
305: $$
306: x\circ y= [[p, x], y]\quad\hbox{  for any } x, y\in \mathfrak{g}_{-1}. 
307: \eqno{(2)}
308: $$
309: Recall that a {\it Jordan algebra} is a commutative
310: algebra $J$ with product $\circ$ satisfying, instead of
311: associativity, the identity
312: $$
313: (x^2\circ y)\circ x = x^2\circ (y\circ x). \eqno{(JI)}
314: $$
315: In a very inspiring paper \cite{Mc} McCrimmon gave an account of some
316: applications of Jordan algebras from antiquity to nowadays, see also
317: refs.  in \cite{U}, \cite{Koe}.  The paper and books strengthen my
318: prejudice that {\it general} Jordan algebras are, bluntly speaking,
319: useless.  Contrariwise, {\it simple} Jordan algebras give rise to
320: several notions important in various problems.  For simple Jordan
321: algebras the so-called {\it general norm} \cite{K4} should be
322: nondegenerate which imposes additional constraints on the parabolic
323: subalgebra.  This answers the above questions but tempts one to make
324: use of the other coset spaces as well.
325: 
326: I wish to make similar use of simple Jordan {\it super}algebras,
327: especially infinite dimensional ones, associated with infinite
328: dimensional classical superdomains listed in \cite{LSV}; for
329: convenience I reproduce these tables.
330: 
331: 
332: In the '60s a remarkable correspondence between Jordan algebras and
333: certain $\Zee$-graded Lie algebras became explicit, cf.  \cite{T},
334: \cite{K1}--\cite{K2} and \cite{Koe}.  Kantor used this correspondence
335: to list simple Jordan algebras (over $\Cee$ and $\Ree$) by the, so
336: far, simplest known method.  He clarified the mysterious relation of
337: Jordan algebras with classical domains and actively studied certain
338: generalizations of Jordan algebras associated with $\Zee$-graded
339: simple Lie algebras of finite depth $d$ (since all of them are of the
340: form $\mathfrak{g}=\mathop{\oplus}\limits_{|i|\leq d}\mathfrak{g}_i$,
341: their {\it length} is equal to $d$).  Supersymmetry, supertwistors
342: etc.  are related with gradings $d>1$ almost without exceptions; so it
343: is interesting to find generalizations of Jordan algebras (or, rather,
344: useful related structures).
345: 
346: Following Freudental and Springer, Kantor generalized products $(2)$
347: to several arguments which is natural for $d>1$.  I suggest,
348: contrariwise, to stick to formula $(2)$, even for $d>1$, with the
349: minimal modification: fix $p\in\mathfrak{g}_{1}$ and for $\mathfrak{
350: g}_{-}=\mathop{\oplus}\limits_{i\leq 0}\mathfrak{g}_i$ set
351: $$
352: x\circ y= [[p, x], y]\quad\hbox{  for any } x, y\in
353: \mathfrak{g}_{-}.  \eqno{(3)}
354: $$
355: In this way we obtain noncommutative generalizations of Jordan
356: algebras (with unknown relations instead of (JI)) and it is
357: interesting to investigate what type of integrable systems are
358: associated with them under the Sokolov-Svinolupov's approach, cf. 
359: \cite{SS}, \cite{HSY}.
360: 
361: Kac \cite{Ka} has already applied this correspondence to 
362: list simple {\it finite} dimensional Jordan superalgebras. 
363: 
364: Tables (borrowed from \cite{LSV}) provide with a list of simple Jordan
365: superalgebras associated with the known in 1991 simple $\Zee$-graded
366: Lie superalgebras of polynomial growth (SZGLSAPGs for short), cf. 
367: \cite{Ka}, \cite{LSh2}, \cite{KMZ}, namely, with $\Zee$-gradings of
368: depth 1 of SZGLSAPGs {\it including} finite dimensional ones.  (This
369: is the place where \cite{Ka} contains an omission --- cf.  Kac'
370: exceptional Jordan superalgebra $K$ with our series $\mathfrak{sh}$
371: discovered by Serganova in 1983, see \cite{L}, and later rediscovered
372: several times.)
373: 
374: \subsubsection{Tits--Kantor--K\"ocher's functor $\mathfrak{kan}$ }
375: 
376: Let $J$ be a Jordan superalgebra and $p$ the tensor that determines
377: the product in $J$, i.e., $p(x, y) = x \circ y$.  To $J$, Kantor
378: assigns (see \cite{Ka}) a $\Zee$-graded Lie superalgebra
379: $\mathfrak{kan}(J)=\mathop{\oplus}\limits_{|i|\leq 1}
380: \mathfrak{kan}(J)_{i}$, a $\Zee$-graded Lie subalgebra in
381: $\mathfrak{vect} (J)$ such that (here $L_a(x)= a\circ x$)
382: $$
383: \renewcommand{\arraystretch}{1.4}
384: \begin{array}{l}
385: \mathfrak{kan}(J)_{-1}=\mathfrak{vect} (J)_{-1}, \\
386: \mathfrak{kan}(J)_{0}= \Span( L_{a}, [L_a, L_b]\mid  a, b \in J),\\ 
387: \mathfrak{kan}(J)_{1}= \Span(p, [L_a, p]\mid  \;a\in J).
388: \end{array}
389: $$
390: 
391: Conversely, for any $\Zee$-graded Lie superalgebra of the form
392: $\mathfrak{g}=\mathop{\oplus}\limits_{i\geq -1}\mathfrak{g}_i$ we
393: define a Jordan superalgebra structure on $\mathfrak{g}_{-1}$ if
394: $(\mathfrak{g}_1)_{\bar{0}} \not=0$ in the following way.  Take $p \in
395: (\mathfrak{g}_1)_{\bar{0}}$ and for $x, y \in \mathfrak{g}_{-1}$ set
396: $$
397: x \circ y = [[p, x], y] . \eqno{(J)}
398: $$
399: 
400: \subsubsection{Digressioin: on homological fields}
401: I do not know what structure is related with an arbitrary odd $p$ but
402: if $p$ is {\it homologic}, i.e., $[p, p]=0$, then the formula
403: $$
404: [x, y]' =[[p, x], y] \eqno{(L)}
405: $$
406: determines a Lie superalgebra structure on $\Pi(\mathfrak{g}_{-1})$. 
407: This structure had been first noticed, perhaps, by M.~Gerstenhaber in
408: '60s and rediscovered many times since then.  It seemed interesting to
409: describe in intrinsic terms the $p$'s which determine {\it simple} Lie
410: (super)algebras.  Homological vector fields were first introduced, in
411: connection with the problem of integration of differential equations
412: on supermanifolds, by V.~Shander \cite{Sha}, who gave a normal form
413: for the nonsingular fields.  However, Shander did not consider
414: singularities of the fields in that work; this was recently done by
415: Vaintrob in a series of articles (e.g.,\cite{V}) in which he showed
416: that the study of singularities of homological fields, and their
417: classification, turns out to be rather similar to the case of
418: singularities of smooth functions.  Regrettably, the answer for $p$
419: corresponding to the simple algebras is more trivial than expected, as
420: we have recently established with Grozman:
421: 
422: Let, first, $\mathfrak{g}=\mathfrak{gl}(n)$; denote the matrix units
423: by $\partial_{i}^j$, let $(\partial_{i}^j)^*=x_{j}^i$ be the dual
424: basis.  Then $p=\sum \xi_{j}^i\xi_{k}^j\delta_{i}^k$, where 
425: $\xi_{j}^i$ is the odd copy of $x_{j}^i$ and $\delta_{i}^k=
426: \frac{\partial}{\partial\xi_{k}^i}$.  Similarly, if
427: the $X_{i}$ form a basis of $\mathfrak{g}$ and $[X_{i}, X_{j}]=\sum
428: c_{ij}^kX_{k}$, then the operator $p\in \mathfrak{vect}(0|\dim
429: \mathfrak{g})$ is of the form $\frac12\sum c_{ij}^kX_{i}^*X_{j}^*
430: X_{k}$, where $X_{i}^*$ is the dual of $X_{i}$.
431: 
432: Having observed that every simple finite dimensional Lie algebra (over
433: $\Cee$) possesses a nondegenerate symmetric bilinear form, we see that
434: for such algebras $p$ is a hamiltonian vector field; to find the
435: corresponding generating function is easy: it is the sum of all
436: elements of degree 1 and weight $0$ with respect to the Cartan
437: subalgebra of $\mathfrak{po}(0|\dim \mathfrak{g})$.  
438: 
439: Generalization to Lie superalgebras is straightforward.  Still,
440: observe that some simple Lie superalgebras have no form at all, some
441: (e.g., $\mathfrak{q}(n)$) possess an {\it odd} nondegenerate symmetric
442: bilinear form in which case $p$ belongs to the antibracket algebra,
443: not to the Poisson one.
444: 
445: \subsection{SZGLSAPGs of depth 1 and length 1} 
446: All possible $\Zee$-gradings of SZGLSAPGs $\mathfrak{g}$ are listed in
447: \cite{Ka} for $\dim \mathfrak{g} < \infty$ and in \cite{L} for most of
448: the other cases.  Our job is to pick those of them which are of depth
449: 1, in particularly, of the form $\mathop{\oplus}\limits_{|i| \leq 1}\;
450: \mathfrak{g}_i$, see Tables.
451: 
452: Albert's notation for Jordan algebras were given in accordance with
453: Cartan's notations for the corresponding Lie algebras.  As follows
454: from Serganova's classification of systems of simple roots of simple
455: Lie superalgebras \cite{S}, Cartan's notations are highly
456: inappropriate for Lie superalgebras.
457: 
458: \subsubsection{Matrix Jordan superalgebras}
459: Let $B_{m, 2n}=\left(\matrix{1_{m}& 0\cr 
460: 0&J_{2n}}\right)$, where 
461: $J_{2n}=\left(\matrix{0&1_{n}\cr -1_{n}&0}\right)$. Set 
462: $$
463: \renewcommand{\arraystretch}{1.4}
464: \begin{array}{l}
465: {\tt Mat}(m|n)=\{X \in \Mat(m|n)\}, \\
466: {\tt Q}(n|n)=
467: \{X \mid [X,J_{2n}]=0\},\\ 
468: {\tt OSp}(m|2n)=
469: \{X \mid X^{st}B_{m, 2n}=B_{m, 2n}X\}, \\ 
470: {\tt Pe}(n|n)=\{X \mid
471: X^{st}J_{2n}=(-1)^{p(X)}J_{2n}X \}. 
472: \end{array}
473: $$
474: In the first two of these spaces the Jordan product is given by the
475: formula
476: $$
477: X \circ Y = XY +(-1)^{p(X)p(Y)}YX.
478: $$
479: I leave it as an excersise to figure out the formula in the other two 
480: cases; for the answer see \cite{Ka}.
481: 
482: \subsubsection{Jordan algebras from bilinear forms} Set
483: $$
484: {\tt Q}_{m|2n}={\Cee}^{m|2n}
485: $$ 
486: with a nondegenerate even symmetric 
487: bilinear form
488: $(\cdot, \cdot)$ and the product
489: $$
490: x\circ y =(e, x)y + x(e, y) -(x, y)e\eqno{(Q)}
491: $$
492: where $e \in ({\tt Q}_{m|2n})_{\bar{0}}$ satisfies $(e, e)=1$.
493: 
494: Set ${\tt HQ}_{m|2n}= \Pi(\Cee [p, q, \Theta])$, where $m>0$, $p=(p_1,
495: \ldots, p_n)$, $q=(q_1, \ldots, q_n)$, $\Theta=(\xi_1, \ldots, \xi_r,
496: \eta_1, \ldots, \eta_r)$ for $m=2r$, of $\Theta=(\xi, \eta, \theta)$
497: for $m=2r+1$ with the Jordan product defined with the help of the
498: symplectic form $\omega$ on the supermanifold with coordinates $p, q,
499: \Theta$:
500: $$
501: x\circ y=\omega (e, x)y + x \omega (e, y) -\omega (x, y)e,\eqno{(H')}
502: $$
503: where $e\in ({\tt HQ}_{m|2n})_{\bar{0}}$ satisfies $\omega (e, e)=1$.
504: 
505: To explicitly give the product, consider the space $\Cee [p, q,
506: \Theta, \alpha, \beta]$ with two extra odd indeterminates and the
507: Poisson bracket such that $p$ and $q$, $\xi$ and $\eta$, and $\alpha,
508: \beta$ are dual.  Setting $\deg \alpha=-\deg \beta=-1$ the degrees of
509: the other indeterminates being $0$, we obtain the $\Zee$-grading of
510: the Poisson algebra, and its quotient modulo center, of the form
511: $(d=1)$.  On $\mathfrak{g}_{-1}= \Cee [p, q, \Theta]\alpha$, define
512: the product \footnotesize
513: $$
514: H_{f\alpha}\circ H_{g\alpha} = \{\{H_{\beta},\,  H_{f\alpha}\}, \, 
515: H_{g\alpha}\}\\
516: = (-1)^{p(f)}H_{\{f, g\}\alpha}.
517: $$ \normalsize
518: In other words, on the superspace of functions with shifted parity, we
519: set
520: $$
521: f\circ g = (-1)^{p(f)+1} \{f, g\}. \eqno{(H,K)}
522: $$
523: 
524: \subsubsection{Exceptional Jordan superalgebras} There are two of them
525: associated with the gradings of $\mathfrak{osp}(4|2; \alpha)$ and
526: $\mathfrak{ab}_3$ from Table 1 and the corresponding loops.
527: 
528: \subsubsection{Stringy Jordan superalgebras} These are obtained from
529: $\mathfrak{k}^L(1|n)$ for $n>2$ and $\mathfrak{k}^M(1|n)$ (see
530: \cite{GLS1}) for $n>3$ by formula $(J)$ with the grading from Table 1. 
531: They will be denoted, respectively, by
532: $$
533: {\tt K{}^LQ}_{1|n} \cong \Pi(\Cee [t^{-1}, t, \theta_1, \dots ,
534: \theta_n])
535: $$
536: $$
537: {\tt K{}^MQ}_{1|n} \cong
538: \Pi(\Cee[t^{-1}, t, \theta_1, \dots , \sqrt{t} \theta_n]).
539: $$
540: The product is given by formula $(H,K)$.
541: 
542:  
543: \subsubsection{Loop Jordan superalgebras} For a finite-dimensional
544: Jordan superalgebra $J$ denote by $J^{(1)}= J \otimes
545: \Cee[t^{-1}, t]$ the loops with values in $J$ and point-wise product.
546: 
547: \subsubsection{Twisted loop Jordan superalgebras} These are associated 
548: by formula $(J)$ with the Lie superalgebras from Table 2.
549: 
550: 	
551: \section{Cross ratios}	
552: In \cite{K3} Kantor generalized the cross ratio of four points on
553: $\Pee^1$ to most of the quotients $G/P$, where $G$ is a simple Lie
554: group and $P$ is its parabolic subgroup.  The Lie algebra
555: $\mathfrak{g}=\hbox{ Lie}(G)$ in these cases is of the form
556: $\mathfrak{g}=\mathop{\oplus}\limits_{|i| \leq d}\; \mathfrak{g}_i$. 
557: I do not know any paper referring to \cite{K3}, so Kantor's studies
558: drew no attention at all.  His constructions, however, naturally
559: appear in supersetting \cite{Mn}; this prompts me to try to decipher a
560: part of \cite{K3}.  I will consider here the simplest case, when $d=1$
561: and the corresponding Jordan algebra is simple.  In this case one can
562: generalize the cross ratio from $\Pee^1={\tt Gr}_{1}^2$ to a
563: collection of ${\tt Gl}(2m|2n)$-invariants of four points on ${\tt
564: Gr}_{m|n}^{2m|2n}$. First, consider \footnotesize
565: $$
566: (A, B, C, D)=(A-B)(C-B)^{-1}(C-D)(A-D)^{-1}.\eqno{(CR)} 
567: $$
568: \normalsize Let $mn=0$.  Now, replace the rhs of (CR) --- call it $X$
569: --- with $\det(X-\lambda E)$. The collection of all coefficients of
570: the powers of $\lambda$ is the analog of the cross ratio.  
571: 
572: By dimension considerations these are all the invariants of four
573: points for the general, orthogonal and Lagrangian grassmannians.
574: 
575: For their super counterparts we take the Berezinian (superdeterminant)
576: and the amount of {\it polynomially independent} invariants is
577: infinite, cf. \cite{Ser}. If, however, we consider rational 
578: dependence, which is natural in super setting, the coefficients of 
579: the first $n+m$ powers of $\lambda$ generate the algebra of 
580: invariants and is a natural candidate for the cross ratio.
581: 
582: On ${\tt Q}(n|n)$, we should take the queerdeterminant, $\qet$, instead of $\det$; the 
583: collection obtained is finite.
584: 
585: For loop Jordan superalgebras we consider matrix-valued functions and $\det(X-\lambda
586: E)$ returns a collection of functions, rather than numbers.  
587: 
588: I do not know the complete cross ratio for Jordan superalgebras
589: related to quadrics and do not know at all what are they for twisted
590: loops and in stringy cases, most interesting to me.  One invariant is
591: obvious (but there should be several if $\dim {\tt Q}>1$): given the form
592: $(\cdot, \cdot)$, or the symplectic form $\omega$ in the curved case,
593: set (for the $\Lambda$-points (see \cite{D}) of the Jordan algebra
594: $J$, i.e., for $A, B, C, D\in (J\otimes \Lambda)_{\bar 0}$
595: \footnotesize
596: $$
597: (A, B, C, D)=\frac{(A-B, A-B)}{(C-B, C-B)}\frac{(C-D, C-D)}{(A-D,
598: A-D)}.\eqno{(CRQ)}
599: $$
600: \normalsize
601: For curved quadrics, take $\omega(H_{A-B}, H_{A-B})$ instead of $(A-B,
602: A-B)$, etc.
603: 
604: Perhaps, other invariants (in non-super case) can be dug out from
605: Reichstein's results.
606: 
607: I almost forgot to add refs.  \cite{EHPSW} that studies four-point
608: functions in $N=2$ superconformal field theories and \cite{M}, where
609: matrix cross ratio is applied to Riccati equation; together with
610: \cite{Mn} they provide a wide setting for applications of our cross
611: ratios.
612: 
613: \section{Tables} Everywhere we assume the notational
614: conventions of \cite{Ser} and definitions adopted there.
615: 
616: In Table 1 we say that the homogeneous superspace $G/P$, where $G$ is a
617: simple Lie supergroup, $P$ its parabolic subsupergroup corresponding to
618: several omitted generators of a Borel subalgebra (description of these
619: generators can be found in \cite{GLP}), of {\it depth} $d$ and {\it
620: length} $l$ if such are the depth and length of $\mathfrak{g}={\rm Lie}(G)$ in the
621: $\Zee$-grading compatible with that of ${\rm Lie}(P)$.  Note that all
622: superspaces of Table 1 possess an hermitian structure (hence are of
623: depth 1) except $PeGr$ (no hermitian structure), $PeQ$ (no hermitian structure,
624: length 2), $CGr_{0, k}^{0, n}$ and $SCGr_{0, k}^{0, n}$ (no hermitian structure,
625: lengths $n-k$ and, resp.,  $n-k-1$).  
626: 
627: Let $\mathfrak{s}(\mathfrak{g})$ be the traceless part of
628: $\mathfrak{g}$ and $\mathfrak{p}(\mathfrak{g})=\mathfrak{g}/{\rm
629: center}$; let $\mathfrak{g}_{\varphi}^{(m)}$ be the stationary
630: subalgebra of the loop algebra with values in $\mathfrak{g}$ singled
631: out by the degree $m$ automorphism $\varphi$ of $\mathfrak{g}$;for
632: $G=\mathfrak{g}_{\varphi}^{(m)}$ with the $\Zee$-grading of type
633: $(d=1)$ the last column of Table 1 contains $G_{0}$; the map $-\st$,
634: ``minus supertransposition'', sends $X$ to $-X^{st}$, the map $\Pi$
635: sends $\left(\matrix{a& b\cr c&d}\right)$ to $\left(\matrix{d& c\cr
636: b&a}\right)$, and $\delta_{x}$ sends $\left(\matrix{a& b\cr c&d}\right)$
637: to $\left(\matrix{a& xb\cr xc&d}\right)$; the automorphism $A$ of 
638: $\mathfrak{po}$ is defined on monomials $f(\theta)$ as $\id$ if 
639: $\frac{\partial f}{\partial \theta_1}=0$ 
640: and otherwise; $irr(\ldots)$ is any of the
641: two irrducible components; $LGr$ and $OGr$ stand for the Lagrangian
642: and orthogonal Grassmannian, respectively; the dual domain is endowed
643: with an asterisk as a left superscript.
644: 
645: Table 2: for the lack of space I give an interpretation of the
646: supergrassmannians here, linewise: the
647: supergrassmannian of $p|q$-dimensional subsuperspaces in $\Cee^{m|n}$
648: and same for $n=m$, $p=q$; superquadric of $1|0$-dimensional isotropic
649: (wrt a non-degenerate even form) lines in $\Cee^{m|n}$; ortolagrangian
650: supergrassmannian; queergrassmannian; ``odd'' superquadric (wrt a
651: non-degenerate even form) of $1|0$-lines in $\Cee^{n|n}$;
652: odd-lagrangian supergrassmannian; curved supergrassmannian of
653: $0|1$-dimensional subsupermanifolds in $\Cee^{0|n}$; curved
654: superquadric; two exceptions.
655: 
656: \begin{table*}[htb]
657: \caption{Gradings of twisted loop 
658: (super)algebras corresponding to hermitian superdomains.}
659: \label{table:1}
660: \tiny
661: $$
662: \renewcommand{\arraystretch}{1.4}
663: \begin{tabular}{cccc}
664: \hline
665: $\mathfrak{g}_{\varphi}^{(m)}$ &$\varphi$ & grading elements from
666: $\mathfrak{h}$& $(\mathfrak{g}_{\varphi}^{(m)})_0$ \cr
667: \hline
668: $\mathfrak{sl}(2m|2n)^{(2)}$&$(-st)\circ \; {\rm Ad}\; {\rm diag}(\Pi
669: _{2m}, J_{2n})$ &${\rm diag}(1_m, -1_m, 1_n,
670: -1_n)$&$\mathfrak{sl}(m|n)^{(1)}$\cr
671: $\mathfrak{sl}(2m)^{(2)}$&$(-t)\circ {\rm Ad}
672: (\Pi_{2m})$&&$\mathfrak{sl}(m)^{(1)}$\cr $\mathfrak{sl}(2n)^{(2)}
673: $&$(t)\circ {\rm Ad} (J _{2n})$&&$\mathfrak{sl}(n)^{(1)}$\cr
674: $\mathfrak{sl}(n|n)^{(2)}$&$\Pi$&${\rm diag}(1_p, 0_{n-p}, 1_p,
675: 0_{n-p})$& $\mathfrak{s}(\mathfrak{gl}(p|p)^{(2)}_{\Pi}\oplus
676: \mathfrak{gl}(n-p|n-p)^{(2)}_{\Pi})$\cr
677: $\mathfrak{sl}(n|n)^{(2)}$&$\Pi\circ (-st)$&${\rm diag}(1_p, -1_{n-p},
678: -1_p, 1_{n-p})$& $\mathfrak{s}(\mathfrak{gl}(p|p)^{(2)}_{\Pi\circ
679: (-st)}\oplus \mathfrak{gl}(n-p|n-p)^{(2)}_{\Pi\circ (-st)})$\cr
680: \hline
681: $\mathfrak{osp}(2m|2n)^{(2)}$&$\varphi_{m, n}{\rm Ad}\; {\rm
682: diag}(1_{2m-1}, 1, 1_{2n})$ &${\rm diag} (J_2, O_{2(m+n-1)}) $&$
683: (\mathfrak{cosp}(2m-2|2n))^{(1)}_{\varphi_{m-1, n}}$\cr
684: $\mathfrak{o}(2m)^{(2)}$& &&$(\mathfrak{co}(2m-2))^{(1)}$\cr
685: \hline
686: $\mathfrak{psq}(2n)^{(4)}$&$ (-st)\circ \delta_{\sqrt{-1}}$&${\rm diag}(J_{2n},
687: J_{2n})$& $\mathfrak{psq}(n)^{(2)}_{\delta_{-1}}$\cr
688: $\mathfrak{sh}(2n)^{(2)}$&$ A$&& $ (\mathfrak{sh}
689: (2n-2)\oplus \Lambda(2n-2))^{(2)}_A $\cr 
690: $\mathfrak{psq}(n)^{(2)}$&$
691: \delta_{-1}$&${\rm diag}(1_p, 0_{n-p}, 1_p, 0_{n-p})$& $\mathfrak{ps}
692: (\mathfrak{q}(p)^{(2)}_{\delta _{-1}}\oplus
693: \mathfrak{q}(n-p)^{(2)}_{\delta _{-1}})$\cr
694: \hline
695: \end{tabular}
696: $$
697: 
698: \end{table*}
699: 
700: \begin{table*}[htb]
701: \caption{Classical superspaces of depth 1.}
702: \label{table:2}
703: \tiny
704: $$
705: \renewcommand{\arraystretch}{1.4}
706: \begin{tabular}{ccccc}
707: \hline
708: $\mathfrak{g} $&$ \mathfrak{g}_0 $&$
709: \mathfrak{g}_{-1}$&Underlying&Name of the\cr 
710: &&&domain&superdomain\cr\hline
711: $\mathfrak{sl}(m|n)$&$ \mathfrak{s}(\mathfrak{gl}(p|q)\oplus
712: \mathfrak{gl}(m-p|n-q))$&$id\otimes id^*$& $Gr_p^m\times
713: Gr_q^n$&$Gr_{p, q}^{m, n}$\cr
714: $\mathfrak{psl}(m|m)$&$\mathfrak{ps}(\mathfrak{gl}(p| p)\oplus
715: \mathfrak{gl}(m-p|m-q))$&$ id\otimes id^*$&$G_p^m\times Gr_p^m
716: $&$Gr_{p, p}^{m, m}$\cr 
717: \hline
718: $\mathfrak{osp}(m| 2n)$&$ \mathfrak{cosp}(m-2| 2n)$& $id$& $Q_{m-2}$&
719: $Q_{m-2, n}$\cr 
720: \hline
721: $\mathfrak{osp}(2m| 2n)$& $\mathfrak{gl}(m|n)$ & $\Lambda^2$&
722:  ${}^*OGr_m\times LGr_n
723: $&$OLGr_{m, n}$ \cr 
724: \hline
725: $\mathfrak{sq}(n)$&$ \mathfrak{s}(\mathfrak{q}(p) \oplus
726: \mathfrak{q}(n-p))$&$irr(id\otimes id*)$& $Gr_p^n$&$ QGr_p^n$\cr
727: $\mathfrak{psq}(n)$&$\mathfrak{ps}(\mathfrak{q}(p) \oplus
728: \mathfrak{q}(n-p))$& &&\cr
729: \hline
730: $\mathfrak{pe}(n)$&$\mathfrak{cpe}(n-1)$&$ id$& $\Cee P^{n-1}$&
731: $PeQ_{n-1}$ \cr 
732: $\mathfrak{spe}(n)$&$\mathfrak{cspe}(n-1)$& &&\cr 
733: \hline
734: $\mathfrak{pe}(n)$&$\mathfrak{gl}(p|n-p)$&$\Pi(S^2 (id))$& $Gr_p^n$&
735: $PeGr_p^n$\cr 
736: $(\mathfrak{spe}(n))$&$ (\mathfrak{sl}(p|n-p))$& or $\Pi
737: (\Lambda^2 (id))$&&\cr
738: \hline
739: \hline
740: $\mathfrak{vect}(0| n)$&$\mathfrak{vect}(0| n-k)\oplus
741: \mathfrak{gl}(k; \Lambda (n-k))$ &$\Lambda (k)\otimes \Pi
742: (id)$&&$CGr_{0, k}^{0, n}$\cr 
743: $\mathfrak{svect}(0|n)$&$\mathfrak{vect}(0| n-k)\oplus \mathfrak{sl}(k; \Lambda (n-k))$&
744: $\Pi (Vol)$ if $k=1$ &--&$SCGr_{0,
745: k}^{0, n}$\cr 
746: \hline
747: $\mathfrak{h}(0| m)$&$\mathfrak{h}(0|m-2)\oplus \Lambda (m-2)\cdot
748: z$&& --&$CQ_{m-2, 0}$ \cr $\mathfrak{sh}(m)$&$\mathfrak{sh}(m-2)\oplus
749: \Lambda (m-2)\cdot z$ &$ \Pi (id)$&&\cr 
750: \hline
751: \hline
752: $\mathfrak{osp}(4|2;\alpha)$&$\mathfrak{cosp}(2|2)\simeq
753: (\mathfrak{gl}(2|1))$ & $id$& $\Cee \Pee^1\times \Cee \Pee^1$ 
754: &$E_{\alpha}$\cr
755: \hline
756: $\mathfrak{ab}(3)$&$\mathfrak{cosp}(2|4)$&$ L_{3\varepsilon_{1}}$ & 
757: $\Cee \Pee^1\times Q_5$ &$AB(3)$\cr \hline
758: \end{tabular}
759: $$\\[2pt]
760: The curved superquadric has infinite dimensional ``stringy'' 
761: counterpart with $\mathfrak{h}(0| m)$ replaced with the centerless 
762: $N$-extended Neveu-Schwarz algebra
763: $\mathfrak{k}^L(1|N)$ or Ramond algebra $\mathfrak{k}^M(1|N)$.
764: \end{table*}
765: 
766: 
767: 
768: \begin{thebibliography}{99}
769: 
770: \bibitem{D}
771: Deligne P. et al (eds.)  {\em Quantum fields and strings: a course for
772: mathematicians}.  Vol.  1, 2.  AMS, Providence, RI, 1999
773: 
774: \bibitem{Sch}
775: I.~Shchepochkina,  Represent.  Theory (electronic
776: journal of AMS), v.  3, 1999, 3 (1999) 373
777: 
778: \bibitem{Ser}
779: A.~Sergeev, Michigan Math.  J. 49 (2001) no.  1, 113 (math.RT/9810113,
780: math.RT/9904079)
781: 
782: 
783: \bibitem{GLS1}
784: P.~Grozman,  D.~Leites and  I.~Shchepochkina, In: M.~Olshanetsky and 
785: A.~Vainshtein (eds.) Multiple facets of quantization and supersymmetry
786: Michael Marinov Memorial Volume, World Sci., to appear [ESI preprint 
787: 1111 (2001)]
788: 
789: \bibitem{LSV} 
790: D.~Leites, V.~Serganova and G.~Vinel. In: C. Bartocci, U. Bruzzo and R.
791: Cianci (eds.)  {\em Differential Geometric Methods in Theoretical
792: Physics} Proc.  DGM-XIX, 1990, Springer, LN Phys.  375 (1991) 286
793: 
794: 
795: \bibitem{Mn}
796: Yu.~Manin, {\em Topics in noncommutative geometry}, Princeton Univ. 
797: Press, Princeton, NJ, 1991 
798: 
799: \bibitem{LSh1} 
800: D.~Leites and I.~Shchepochkina  Classification of simple Lie
801: superalgebras of vector fields, to appear
802: 
803: \bibitem{L}
804: D.~Leites (ed.), {\em Seminar on supermanifolds}, vv.  1--34,
805: 1986--90, Reports of Dept.  of Math.  University of Stockholm, 2100
806: pp.
807: 
808: \bibitem{BL1}
809: J.~Bernstein and D.~Leites, Functional Anal. 
810: Appl.  11 (1977) no.  1, 45
811: 
812: \bibitem{Sh}
813: V.~Shander, Funct.  Anal.  Appl.  22 (1988) no.  1, 80
814: 
815: \bibitem{R}
816: A.~Rudakov, Math.  USSR Izvestiya, v.  38 (1974) n.  4, 835
817: 
818: \bibitem{AS}
819: A.~Shvarts, Nuclear Phys.  B 171 (1980), no.  1-2, 154; A.~Gaiduk,
820: V.~Romanov and A.~Shvarts, Comm.  Math.  Phys.  79 (1981), no.  4,
821: 507; A.~Gaiduk, O.~Khudaverdyan and A.~ Shvarts, Theoret.  and Math. 
822: Phys.  52 (1982), no.  3, 862
823: 
824: \bibitem{BL3}
825: J.~Bernstein and D.~Leites, Selecta Math.  Soviet.  1 (1981) no.  2,
826: 143
827: 
828: \bibitem{BL2}
829: J.~Bernstein and D.~Leites, Functional Analysis and Its Applications
830: 11 (1977) 219
831: 
832: \bibitem{H}
833: O.~Khudaverdian, Comm.  Math.  Phys.  198 (1998), no.  3, 591
834: 
835: \bibitem{LKV}
836: D.~Leites, Yu.~Kochetkov and A.~Vaintrob, In: S.~Andima et.  al (eds.) 
837: {\em General topology and its applications}, LN in pure and applied
838: math, v.  134, Marcel Decker, NY (1991) 217
839: 
840: \bibitem{LSh2}
841: D.~Leites and I.~Shchepochkina, Theor.  and Math.  Physics, v.  126
842: (2000) no.  3, 339
843: 
844: \bibitem{St} 
845: H.~Strade and R.~Farnsteiner, Modular Lie algebras and their
846: representations.  Marcel Dekker, NY, 1988
847: 
848: \bibitem{GLS2}
849: P.~Grozman, D.~Leites and I.~Shchepochkina, hep-th 9702120; Acta
850: Mathematica Vietnamica, v.  26 (2001) no.  1, 27
851: 
852: \bibitem{S2}
853: V.~Serganova, Functional Anal. Appl. 17 (1983), no. 3, 200
854: 
855: \bibitem{Ma}
856: Yu.~Manin, Gauge field theory and complex geometry.  Second edition. 
857: Springer-Verlag, Berlin, 1997
858: 
859: \bibitem{GL}
860: P.~Grozman and D.~Leites In: J.~Wess, E.~Ivanov (eds.), {\em
861: Supersymmetries and quantum symmetries}, (SQS'97, 22--26 July, 1997),
862: Lecture Notes in Phys., 524, 1999, 58
863: 
864: \bibitem{Mc}
865: K.~McCrimmon, Bull.  Amer.  Math.  Soc., v.84 (1978) no.  4, 612
866: 
867: \bibitem{U}
868: H.~Upmeier,Jordan algebras in analysis, operator theory, and quantum
869: mechanics.  CBMS Regional Conference Series in Mathematics, 67.  AMS,
870: Providence, RI, 1987.
871: 
872: \bibitem{Koe}
873: M.~Koecher (edited by A.~Krieg and S.~Walcher) {\em The Minnesota
874: Notes on Jordan Algebras and Their Applications}, Lect.  Notes Math.,
875: 1710, Springer, 1999
876: 
877: \bibitem{K4}
878: I.~Kantor, Proc.  of P.~K.~Rashevsky's seminar on vector and tensor
879: analysis.  v. 14 (1968), 114 (in Russian)
880: 
881: \bibitem{T}
882: J.~Tits, Indag. Math. v. 24 (1962) 530
883: 
884: \bibitem{K1}
885: I.~Kantor, Soviet. Math. Doklady v. 5 (1964) 1404
886: 
887: \bibitem{K2}
888: I.~Kantor, Proc.  of P.~K.~Rashevsky's seminar on vector and tensor
889: analysis.  v. 13 (1966), 310 (in Russian)
890: 
891: \bibitem{SS}
892: S.~Svinolupov and V.~Sokolov, Theoret.  and Math.  Phys.  108 (1996)
893: no.  3, 1160; id., Acta Appl.  Math.  41 (1995) no.  1--3, 323; id.,
894: Math.  Notes 53 (1993) no.  1--2, 201
895: 
896: \bibitem{HSY}
897: I.~Habibullin, V.~Sokolov and R.~Yamilov.  In: Nonlinear physics:
898: theory and experiment (Lecce, 1995) World Sci., River Edge, NJ, 1996,
899: 139
900: 
901: \bibitem{Ka}
902: V.~Kac, Commun.  Alg. 5 (13) (1977) 1375; L.~Hogben and V.~Kac,
903: Erratum Comm.  Algebra 11 (1983) no.  10, 1155
904: 
905: \bibitem{KMZ}
906: V.~Kac, C.~Martinez and E.~Zelmanov, Graded simple Jordan
907: superalgebras of growth one.  Mem.  Amer.  Math.  Soc.  150 (2001)
908: no.  711
909: 
910: \bibitem{Sha}
911: V.~Shander, Functional Anal. Appl. 14 (1980), no. 2, 160
912: 
913: \bibitem{V}
914: A.~Vaintrob, J. Math.  Sci.  82 (1996), no.  6, 3865
915: 
916: \bibitem{S}
917: V.~Serganova, Comm.  Algebra 24 (1996) no.  13, 4281
918: 
919: \bibitem{K3}
920: I.~Kantor, Proc.  of P.~K.~Rashevsky's seminar on vector and tensor
921: analysis.  v. 17 (1974) 250; v. 18 (1975) 234 (in Russian)
922: 
923: 
924: \bibitem{GLP}
925: P.~Grozman, D.~Leites and E.~Poletaeva,
926: In: E.~Ivanov et.  al.  (eds.)  {\em Supersymmetries and Quantum
927: Symmetries} (SQS'99, 27--31 July, 1999), Dubna, JINR, 2000, 387
928: 
929: \bibitem{EHPSW}
930: B.~Eden, P.~Howe, A.~Pickering,
931: E.~Sokatchev, P.~West, Nuclear Phys.  B 581 (2000), no.  1-2, 523--558
932: 
933: \bibitem{M}
934: M.~Zelikin,{\em Control theory and optimization.  I.
935: Homogeneous spaces and the Riccati equation in the calculus of
936: variations}. Encyclopaedia of Mathematical
937: Sciences, 86.  Springer-Verlag, Berlin, 2000.  xii+284 pp.
938: 
939: \end{thebibliography}
940: 
941: \end{document}
942: