-
Notifications
You must be signed in to change notification settings - Fork 1
Expand file tree
/
Copy pathcatsem.tex
More file actions
382 lines (320 loc) · 16.1 KB
/
Copy pathcatsem.tex
File metadata and controls
382 lines (320 loc) · 16.1 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
\chapter{Categorical semantics}\label{categorical-semantics}
Constructing denotational models of linear logic can be a tedious work.
Categorical semantics are useful to identify the fundamental structure
of these models, and thus simplify and make more abstract the
elaboration of those models.
% \begin{quotation}
% \texttt{TODO: why categories? how to extract categorical models? etc.}
% \end{quotation}
See~\cite{categoriesworkmath} for a more detailed introduction to category theory. See~\cite{catsemll} for a detailed treatment of categorical semantics of linear logic.
% \section{Basic category theory recalled}\label{basic-category-theory-recalled}
% \begin{definition}[Category]
% \end{definition}
% \begin{definition}[Functor]
% \end{definition}
% \begin{definition}[Natural transformation]
% \end{definition}
% \begin{definition}[Adjunction]
% \end{definition}
% \begin{definition}[Monad]
% \end{definition}
\section{Overview}\label{overview}
In order to interpret the various \hyperref[fragment]{fragments} of linear
logic, we define incrementally what structure we need in a categorical
setting.
\begin{itemize}
\item The most basic underlying structure are \emph{symmetric monoidal
categories} which model the symmetric tensor \(\otimes\) and its unit
\(1\).
\item The \(\otimes, \multimap\) fragment (\hyperref[imll]{IMLL}) is captured by
so-called \emph{symmetric monoidal closed categories}.
\item Upgrading to \hyperref[intuitionistic-linear-logic]{ILL}, that is, adding the exponential \(\oc\)
modality to IMLL requires modelling it categorically. There are
various ways to do so: using rich enough \emph{adjunctions}, or with
an ad-hoc definition of a well-behaved comonad which leads to
\emph{linear categories} and close relatives.
\item Dealing with the additives \(\with, \oplus\) is quite easy, as they
are plain \emph{cartesian product} and \emph{coproduct}, usually
defined through universal properties in category theory.
\item Retrieving \(\parr\), \(\bot\) and \(\wn\) is just a matter of
dualizing \(\otimes\), \(1\) and \(\oc\), thus requiring the model to
be a \emph{*-autonomous category} for that purpose.
\end{itemize}
\section{\texorpdfstring{Modeling \hyperref[imll]{IMLL}}{Modeling IMLL}}\label{modeling-imll}
A model of \hyperref[imll]{IMLL} is a \emph{closed symmetric monoidal category}. We
recall the definition of these categories below.
\begin{definition}[Monoidal category]\label{monoidalcategory}
A \emph{monoidal category} $(\mathcal{C},\otimes,I,\alpha,\lambda,\rho)$ is a category $\mathcal{C}$ equipped with
\begin{itemize}
\item a functor $\otimes:\mathcal{C}\times\mathcal{C}\to\mathcal{C}$ called \emph{tensor product},
\item an object $I$ called \emph{unit object},
\item three natural isomorphisms $\alpha$, $\lambda$ and $\rho$, called respectively \emph{associator}, \emph{left unitor} and \emph{right unitor}, whose components are
\begin{equation*}
\alpha_{A,B,C}:(A\otimes B)\otimes C\to A\otimes (B\otimes C)
\qquad
\lambda_A:I\otimes A\to A
\qquad
\rho_A:A\otimes I\to A
\end{equation*}
\end{itemize}
such that
\begin{itemize}
\item for every objects $A,B,C,D$ in $\mathcal{C}$, the diagram
\begin{equation*}
\xymatrix{
((A\otimes B)\otimes C)\otimes D\ar[d]_{\alpha_{A\otimes B,C,D}}\ar[r]^{\alpha_{A,B,C}\otimes D}& (A\otimes(B\otimes C))\otimes D\ar[r]^{\alpha_{A,B\otimes C,D}}& A\otimes((B\otimes C)\otimes D)\ar[d]^{A\otimes\alpha_{B,C,D}}\\
(A\otimes B)\otimes(C\otimes D)\ar[rr]_{\alpha_{A,B,C\otimes D}}& & A\otimes(B\otimes (C\otimes D))
}
\end{equation*}
commutes,
\item for every objects $A$ and $B$ in $\mathcal{C}$, the diagram
\begin{equation*}
\xymatrix{
(A\otimes I)\otimes B\ar[dr]_{\rho_A\otimes B}\ar[rr]^{\alpha_{A,I,B}}& & \ar[dl]^{A\otimes\lambda_B}A\otimes(I\otimes B)\\
& A\otimes B&
}
\end{equation*}
commutes.
\end{itemize}
\end{definition}
\begin{definition}[Braided, symmetric monoidal category]
A \emph{braided} monoidal category is a category together with a natural isomorphism of components
\begin{equation*}
\gamma_{A,B}:A\otimes B\to B\otimes A
\end{equation*}
called \emph{braiding}, such that the two diagrams
\begin{equation*}
\xymatrix{
& A\otimes(B\otimes C)\ar[r]^{\gamma_{A,B\otimes C}}& (B\otimes C)\otimes A\ar[dr]^{\alpha_{B,C,A}}\\
(A\otimes B)\otimes C\ar[ur]^{\alpha_{A,B,C}}\ar[dr]_{\gamma_{A,B}\otimes C}& & & B\otimes (C\otimes A)\\
& (B\otimes A)\otimes C\ar[r]_{\alpha_{B,A,C}}& B\otimes(A\otimes C)\ar[ur]_{B\otimes\gamma_{A,C}}\\
}
\end{equation*}
and
\begin{equation*}
\xymatrix{
& (A\otimes B)\otimes C\ar[r]^{\gamma_{A\otimes B,C}}& C\otimes (A\otimes B)\ar[dr]^{\alpha^{-1}_{C,A,B}}& \\
A\otimes (B\otimes C)\ar[ur]^{\alpha^{-1}_{A,B,C}}\ar[dr]_{A\otimes\gamma_{B,C}}& & & (C\otimes A)\otimes B\\
& A\otimes(C\otimes B)\ar[r]_{\alpha^{-1}_{A,C,B}}& (A\otimes C)\otimes B\ar[ur]_{\gamma_{A,C}\otimes B}& \\
}
\end{equation*}
commute for every objects $A$, $B$ and $C$.
A \emph{symmetric} monoidal category is a braided monoidal category in which the braiding satisfies
\begin{equation*}
\gamma_{B,A}\circ\gamma_{A,B}=A\otimes B
\end{equation*}
for every objects $A$ and $B$.
\end{definition}
\begin{definition}[Closed monoidal category]\label{smcc}
A monoidal category $(\mathcal{C},\tens,I)$ is \emph{left closed} when for every object $A$, the functor $B\mapsto A\otimes B$ has a right adjoint, written $B\mapsto(A\limp B)$.
This means that there exists a bijection
\begin{equation*}
\mathcal{C}(A\tens B, C) \cong \mathcal{C}(B,A\limp C)
\end{equation*}
which is natural in $B$ and $C$.
Equivalently, a monoidal category is left closed when it is equipped with a \emph{left closed structure}, which consists of
\begin{itemize}
\item an object $A\limp B$,
\item a morphism $\mathrm{eval}_{A,B}:A\tens (A\limp B)\to B$, called \emph{left evaluation},
\end{itemize}
for every objects $A$ and $B$, such that for every morphism $f:A\otimes X\to B$ there exists a unique morphism $h:X\to A\limp B$ making the diagram
\begin{equation*}
\xymatrix{
A\tens X\ar@{.>}[d]_{A\tens h}\ar[dr]^{f}\\
A\tens(A\limp B)\ar[r]_-{\mathrm{eval}_{A,B}}& B
}
\end{equation*}
commute.
Dually, the monoidal category $\mathcal{C}$ is \emph{right closed} when the functor $B\mapsto B\otimes A$ admits a right adjoint. The notion of \emph{right closed structure} can be defined similarly.
\end{definition}
In a symmetric monoidal category, a left closed structure induces a
right closed structure and conversely, allowing us to simply speak of a
\emph{closed symmetric monoidal category}.
\section{Modeling the additives}\label{modeling-the-additives}
\begin{definition}[Product]\label{cartesianproduct}
A \emph{product} $(X,\pi_1,\pi_2)$ of two coinitial morphisms $f:A\to B$ and $g:A\to C$ in a category $\mathcal{C}$ is an object $X$ of $\mathcal{C}$ together with two morphisms $\pi_1:X\to B$ and $\pi_2:X\to C$ such that there exists a unique morphism $h:A\to X$ making the diagram
\begin{equation*}
\xymatrix{
& \ar[ddl]_fA\ar@{.>}[d]_h\ar[ddr]^g& \\
& \ar[dl]^{\pi_1}X\ar[dr]_{\pi_2}& \\
B& & C
}
\end{equation*}
commute.
$(X,\pi_1,\pi_2)$ is a product of the two objects $B$ and $C$ if it is a product of any two coinitial morphisms $f:A\to B$ and $g:A\to C$.
\end{definition}
A category has \emph{finite products} when it has products and a terminal object.
\begin{definition}[Monoid]
A \emph{monoid} $(M,\mu,\eta)$ in a monoidal category $(\mathcal{C},\tens,I)$ is an object $M$ together with two morphisms $\mu:M\tens M \to M$ and $\eta:I\to M$
such that the diagrams
\begin{equation*}
\xymatrix{
& (M\tens M)\tens M\ar[dl]_{\alpha_{M,M,M}}\ar[r]^-{\mu\tens M}& M\tens M\ar[dd]^{\mu}\\
M\tens(M\tens M)\ar[d]_{M\tens\mu}& & \\
M\tens M\ar[rr]_{\mu}& & M\\
}
\end{equation*}
and
\begin{equation*}
\xymatrix{
I\tens M\ar[r]^{\eta\tens M}\ar[dr]_{\lambda_M}& M\tens M\ar[d]_\mu& \ar[l]_{M\tens\eta}\ar[dl]^{\rho_M}M\tens I\\
& M&
}
\end{equation*}
commute.
\end{definition}
\begin{property}
Categories with products are monoidal categories.
\end{property}
\section{\texorpdfstring{Modeling \hyperref[intuitionistic-linear-logic]{ILL}}{Modeling ILL}}\label{modeling-ill}
\begin{definition}[Linear-non linear (LNL) adjunction~\cite{mixedlnll}]
A \emph{linear-non linear adjunction} is a symmetric monoidal adjunction between lax monoidal functors
\begin{equation*}
\xymatrix{
(\mathcal{M},\times,\top)\ar@/^/[rr]^{(L,l)}& \bot& \ar@/^/[ll]^{(M,m)}(\mathcal{L},\otimes,I)
}
\end{equation*}
in which the category $\mathcal{M}$ has finite products.
\end{definition}
\[\oc=L\circ M\]
This section is devoted to defining the concepts necessary to define
these adjunctions.
\begin{definition}[Monoidal functor]
A \emph{lax monoidal functor} $(F,f)$ between two monoidal categories $(\mathcal{C},\tens,I)$ and $(\mathcal{D},\bullet,J)$ consists of
\begin{itemize}
\item a functor $F:\mathcal{C}\to\mathcal{D}$ between the underlying categories,
\item a natural transformation $f$ of components $f_{A,B}:FA\bullet FB\to F(A\tens B)$,
\item a morphism $f:J\to FI$
\end{itemize}
such that the diagrams
\begin{equation*}
\xymatrix{
(FA\bullet FB)\bullet FC\ar[d]_{\phi_{A,B}\bullet FC}\ar[r]^{\alpha_{FA,FB,FC}}& FA\bullet(FB\bullet FC)\ar[dr]^{FA\bullet\phi_{B,C}}\\
F(A\otimes B)\bullet FC\ar[dr]_{\phi_{A\otimes B,C}}& & FA\bullet F(B\otimes C)\ar[d]^{\phi_{A,B\otimes C}}\\
& F((A\otimes B)\otimes C)\ar[r]_{F\alpha_{A,B,C}}& F(A\otimes(B\otimes C))
}
\end{equation*}
and
\begin{equation*}
\vcenter{\vbox{\xymatrix{
FA\bullet J\ar[d]_{\rho_{FA}}\ar[r]^{FA\bullet\phi}& FA\bullet FI\ar[d]^{\phi_{A,I}}\\
FA& \ar[l]^{F\rho_A}F(A\otimes I)
}}}
\qquad\text{and}\qquad
\vcenter{\vbox{\xymatrix{
J\bullet FB\ar[d]_{\lambda_{FB}}\ar[r]^{\phi\bullet FB}& FI\bullet FB\ar[d]^{\phi_{I,B}}\\
FB& \ar[l]^{F\lambda_B}F(I\otimes B)
}}}
\end{equation*}
commute for every objects $A$, $B$ and $C$ of $\mathcal{C}$. The morphisms $f_{A,B}$ and $f$ are called \emph{coherence maps}.
A lax monoidal functor is \emph{strong} when the coherence maps are invertible and \emph{strict} when they are identities.
\end{definition}
\begin{definition}[Monoidal natural transformation]
Suppose that $(\mathcal{C},\tens,I)$ and $(\mathcal{D},\bullet,J)$ are two monoidal categories and
$(F,f):(\mathcal{C},\tens,I)\Rightarrow(\mathcal{D},\bullet,J)$
and
$(G,g):(\mathcal{C},\tens,I)\Rightarrow(\mathcal{D},\bullet,J)$
are two monoidal functors between these categories. A \emph{monoidal natural transformation} $\theta:(F,f)\to (G,g)$ between these monoidal functors is a natural transformation $\theta:F\Rightarrow G$ between the underlying functors such that the diagrams
\begin{equation*}
\vcenter{\vbox{\xymatrix{
FA\bullet FB\ar[d]_{f_{A,B}}\ar[r]^{\theta_A\bullet\theta_B}& \ar[d]^{g_{A,B}}GA\bullet GB\\
F(A\tens B)\ar[r]_{\theta_{A\tens B}}& G(A\tens B)
}}}
\qquad\text{and}\qquad
\vcenter{\vbox{\xymatrix{
& \ar[dl]_{f}J\ar[dr]^{g}& \\
FI\ar[rr]_{\theta_I}& & GI
}}}
\end{equation*}
commute for every objects $A$ and $B$ of $\mathcal{D}$.
\end{definition}
\begin{definition}[Monoidal adjunction]
A \emph{monoidal adjunction} between two monoidal functors
$(F,f):(\mathcal{C},\tens,I)\Rightarrow(\mathcal{D},\bullet,J)$
and
$(G,g):(\mathcal{D},\bullet,J)\Rightarrow(\mathcal{C},\tens,I)$
is an adjunction between the underlying functors $F$ and $G$ such that the unit and the counit
$\eta:\mathcal{C}\Rightarrow G\circ F$ and $\varepsilon:F\circ G\Rightarrow\mathcal{D}$
induce monoidal natural transformations between the corresponding monoidal functors.
\end{definition}
\section{Modeling negation}\label{modeling-negation}
\subsection{*-autonomous categories}\label{autonomous-categories}
\begin{definition}[*-autonomous category]\label{staraut}
Suppose that we are given a symmetric monoidal closed category $(\mathcal{C},\tens,I)$ and an object $R$ of $\mathcal{C}$. For every object $A$, we define a morphism $\partial_{A}:A\to(A\limp R)\limp R$ as follows. By applying the bijection of the adjunction defining (left) closed monoidal categories to the identity morphism $\mathrm{id}_{A\limp R}:A\limp R \to A\limp R$, we get a morphism $A\tens (A\limp R)\to R$, and thus a morphism $(A\limp R)\tens A\to R$ by precomposing with the symmetry $\gamma_{A\limp R,A}$. The morphism $\partial_A$ is finally obtained by applying the bijection of the adjunction defining (left) closed monoidal categories to this morphism. The object $R$ is called \emph{dualizing} when the morphism $\partial_A$ is a bijection for every object $A$ of $\mathcal{C}$. A symmetric monoidal closed category is \emph{*-autonomous} when it admits such a dualizing object.
\end{definition}
\subsection{Compact closed categories}\label{compact-closed-categories}
\begin{definition}[Dual objects]
A \emph{dual object} structure $(A,B,\eta,\varepsilon)$ in a monoidal category $(\mathcal{C},\tens,I)$ is a pair of objects $A$ and $B$ together with two morphisms
$\eta:I\to B\otimes A$ and $\varepsilon:A\otimes B\to I$
such that the diagrams
\begin{equation*}
\xymatrix{
& A\tens(B\tens A)\ar[r]^{\alpha_{A,B,A}^{-1}}& (A\tens B)\tens A\ar[dr]^{\varepsilon\tens A}\\
A\tens I\ar[ur]^{A\tens\eta}& & & I\tens A\ar[d]^{\lambda_A}\\
A\ar[u]^{\rho_A^{-1}}\ar@{=}[rrr]& & & A\\
}
\end{equation*}
and
\begin{equation*}
\xymatrix{
& (B\tens A)\tens B\ar[r]^{\alpha_{B,A,B}}& B\tens(A\tens B)\ar[dr]^{B\tens\varepsilon}\\
I\tens B\ar[ur]^{\eta\tens B}& & & B\tens I\ar[d]^{\rho_B}\\
B\ar[u]^{\lambda_B^{-1}}\ar@{=}[rrr]& & & B\\
}
\end{equation*}
commute. The object $A$ is called a left dual of $B$ (and conversely $B$ is a right dual of $A$).
\end{definition}
\begin{lemma}
Two left (resp.\ right) duals of a same object $B$ are necessarily isomorphic.
\end{lemma}
\begin{definition}[Compact closed category]
A symmetric monoidal category $(\mathcal{C},\tens,I)$ is \emph{compact closed} when every object $A$ has a right dual $A^*$. We write
$\eta_A:I\to A^*\tens A$ and $\varepsilon_A:A\tens A^*\to I$
for the corresponding duality morphisms.
\end{definition}
\begin{lemma}
In a compact closed category the left and right duals of an object $A$ are isomorphic.
\end{lemma}
\begin{property}
A compact closed category $\mathcal{C}$ is monoidal closed, with closure defined by
$\mathcal{C}(A\tens B,C)\cong\mathcal{C}(B,A^*\tens C)$.
\end{property}
\begin{proof}
To every morphism $f:A\tens B\to C$, we associate a morphism $\ulcorner f\urcorner:B\to A^*\tens C$ defined as
\begin{equation*}
\xymatrix{
B\ar[r]^-{\lambda_B^{-1}}& I\tens B\ar[r]^-{\eta_A\tens B}& (A^*\tens A)\tens B\ar[r]^-{\alpha_{A^*,A,B}}& A^*\tens(A\tens B)\ar[r]^-{A^*\tens f}& A\tens C\\
}
\end{equation*}
and to every morphism $g:B\to A^*\tens C$, we associate a morphism $\llcorner g\lrcorner:A\tens B\to C$ defined as
\begin{equation*}
\xymatrix{
A\tens B\ar[r]^-{A\tens g}& A\tens(A^*\tens C)\ar[r]^-{\alpha_{A,A^*,C}^{-1}}& (A\tens A^*)\tens C\ar[r]^-{\varepsilon_A\tens C}& I\tens C\ar[r]^-{\lambda_C}& C
}
\end{equation*}
It is easy to show that $\llcorner \ulcorner f\urcorner\lrcorner=f$ and $\ulcorner\llcorner g\lrcorner\urcorner=g$ from which we deduce the required bijection.
\end{proof}
\begin{property}
A compact closed category is a (degenerated) *-autonomous category, with the obvious duality structure. In particular, $(A \otimes B)^* \cong A^*\otimes B^*$.
\end{property}
\begin{remark}
The above isomorphism does not hold in *-autonomous categories in general. This means that models which are compact closed categories identify $\otimes$ and $\parr$ as well as $1$ and $\bot$.
\end{remark}
\begin{proof}
The dualizing object $R$ is simply $I^*$.
For any $A$, the reverse isomorphism $\delta_A : (A \multimap R)\multimap R \rightarrow A$ is constructed as follows:
\begin{equation*}
\mathcal{C}((A \multimap R)\multimap R, A) := \mathcal{C}((A \otimes I^{**})\otimes I^{**}, A) \cong \mathcal{C}((A \otimes I)\otimes I, A) \cong \mathcal{C}(A, A)
\end{equation*}
Identity on $A$ is taken as the canonical morphism required.
\end{proof}
% \section{Other categorical models}\label{other-categorical-models}
% \subsection{Lafont categories}\label{lafont-categories}
% \subsection{Seely categories}\label{seely-categories}
% \subsection{Linear categories}\label{linear-categories}
% \section{Properties of categorical models}\label{properties-of-categorical-models}
% \subsection{The Kleisli category}\label{the-kleisli-category}
%%% Local Variables:
%%% mode: latex
%%% TeX-master: "main"
%%% End: