• Aucun résultat trouvé

On the homological dimensions of pullbacks

N/A
N/A
Protected

Academic year: 2022

Partager "On the homological dimensions of pullbacks"

Copied!
5
0
0

Texte intégral

(1)

THÉORIE DES NOMBRES BESANÇON

Armées 1998/2001

O n t h e h o m o l o g i c a l d i m e n s i o n s o f p u l l b a c k s . I I

N . K O S M A T O V

(2)

O n t h e h o m o l o g i c a l d i m e n s i o n s o f p u l l b a c k s . I I

N i k o l a i K o s m a t o v

In this article, ail rings are assumed t o have identity elements preserved by ring h o m o m o r p h i s m s , a n d ail modules are left modules. For a ring À, let lgld A a n d w d A d é n o t é t h e left global dimension of A and t h e weak dimension of A, respectively. For a A-module X , we d é n o t é t h e injective, projective a n d flat dimensions of X by idA X , p dA X a n d fdA X , respectively.

T h e left finitistic injective, projective a n d flat dimensions of A are denoted a n d deflned as follows:

1FID A = sup{idA M | M is a A-module w i t h idA M < oo}, 1FPD A = s u p { p dA M | M is a A-module with p dA M < oo},

1FFD A = s u p { f dA M | M is a A-module with fdA M < oo}.

Consider a c o m m u t a t i v e square of rings a n d ring h o m o m o r p h i s m s R —ii-»- Rx

h ii (1)

R2 R',

where R is t h e pullback (also called fibre p r o d u c t ) of R i a n d R2 over R', t h a t is, given r} G R i a n d r2 G R2 w i t h j i ( r i ) = j2( r2) , there is a unique element r G R such t h a t i \ ( r ) = rx and i2( r ) = r2. W e assume t h a t is a surjection.

In t h e present p a p e r we continue our s t u d y of homological dimensions of pullbacks s t a r t e d in [1]. O u r p u r p o s e is t o give u p p e r b o u n d s for t h e finitistic dimensions of R ( T h e o r e m s 1, 2 a n d 3). W e also provide two simple examples of pullbacks where we use these results t o calculate homological dimensions, a n d show t h a t our conditions are essential. In t h e first example, a pullback of two h e r e d i t a r y rings has finite finitistic dimensions t h o u g h its global and weak dimensions are infinité. Therefore, it is impossible t o e s t i m a t e t h e global a n d weak dimensions of a pullback if only t h a t of t h e c o m p o n e n t rings are given. T h e second example d e m o n s t r a t e s t h a t our e s t i m â t e s would not be t r u e if we d r o p p e d t h e a s s u m p t i o n t h a t i\ is surjective.

1

(3)

T h e o r e m 1. Let n be a non-negative integer. Suppose that f o r every R.- module M of finite injective dimension we have that

i d / ^ E x t l R(Rk, M ) ) ^ n - l f o r l = 0, 1, . . . , n a n d k = 1 , 2 . Then 1FID R ^ n .

P r o o f . Let M b e an i2-module of finite injective dimension.

F r o m [1, Proposition 5] it follows t h a t i dRM ^ n. Therefore 1FID R = s u p { i df l M \ i dH M < oo} ^ n.

Similarly, [1, Propositions 6 a n d 7] allow us t o prove analogous b o u n d s for finitistic projective a n d fiât dimensions.

T h e o r e m 2 . Let n be a non-negative integer. Suppose that f o r every R - module M of finite projective dimension we have that

p dH f c( T o r f ( Rk, M ) ) ^ n - l f o r 1 = 0, 1, . . . , n a n d k = 1 , 2 . Then 1FPD R ^ n .

T h e o r e m 3 . Let n be a non-negative integer. Suppose that f o r every R - module M of finite f l a t dimension we have that

f d ^ ( T o r f ( Rk, M ) ) ^ n - l f o r l = 0, 1, . . . , n a n d k = l , 2 . Then 1FFD R < n .

E x a m p l e 1. Let s ^ 2, R ' = Z / s Z , R i = R2 = Z, R = { ( m i , m2) e R i x R2\ m i = m2 (mod s) }. T h e n in t h e c o m m u t a t i v e square (1) w i t h canonical surjections ik a n d jk t h e ring R is t h e pullback of R \ a n d R2 over R'. T h e r e exist t h e periodic free resolutions of t h e f?-modules Rk

. . . H R m R H R - ^ R x - ï 0, (2)

(a,0) nR ^ R ^ R R (0,3) 0 fs,o; 2 0, (3)

where t h e syzygies are t h e s u b m o d u l e s s Z x 0 ~ R i a n d 0 x s Z ~ R2. It is easily seen t h a t t h e short exact sequences

0 — • 0 x s Z <-> R R i — • 0,

(4)

do not split. Hence t h e Zï-modules Rk are not projective. B y [2, T h e o r e m 3.2.7], t h e y are not fiât either. It follows t h a t p dR Rk = f d # Rk = oo and l g l d i ? = w d i ? = oo. At t h e same time, l g l d Rk = wd Rk — 1. W e see t h a t it is impossible t o e s t i m a t e lgld R a n d wd R with only lgld Rk a n d wd Rk given.

Let M b e a n /^-module of finite projective dimension w i t h a projective resolution

. . . — > 0 — • 0 — » Pn — * . . . — • P0 — • M — » 0. (4) Since l g l d Rk = 1, we have pdf l f c(Toro ( Rk, M ) ) < 1. A p p l y i n g [2, Exercise

2.4.3] t o t h e projective resolutions (2), (3) a n d (4), we o b t a i n for a sufficiently large t

T o r f ( Rk, M ) ~ T o v f+ 2 t( Rk, M ) ~ T o r f ( i ?f c, 0 ) = 0.

Consequently, p df l f c( T o r f ( Rk, M ) ) = pdHfc 0 = 0. T h e o r e m 2 now yields t h a t 1FPD R ^ 1. In t h e s a m e m a n n e r we can use T h e o r e m s 1 a n d 3 t o show t h a t 1FID R ^ l a n d l F F D i ? ^ 1.

Consider t h e following projective resolution of t h e -R-module R / ( s , s ) R : 0 — • R R - A R / ( s , s ) R — 0 .

Since this short exact sequence does not split, we have p dr( R / ( s , s ) R ) = f dr( R / ( s , s ) R ) = 1. T h i s clearly forces 1FPD R = 1FFD R = 1.

T h e s u b g r o u p R of t h e free Abelian g r o u p Z x Z is a free Abelian group also, therefore, applying t h e f u n c t o r H o m z (R, — ) t o t h e short exact sequence 0 —>• Z Q —> Q / Z 0, we o b t a i n a short exact sequence of i?-modules

0 — • H o mz( i ? , Z ) —> H o mz(JR , Q ) —•> H o mz( i ? , Q / Z ) —> 0.

T h i s sequence does not split and, by [2, Corollary 2.3.11], it is an injective resolution of t h e i î - m o d u l e H o m z ( i î , Z). It follows t h a t id/î(Homz(.R, Z)) = 1, a n d hence t h a t 1FID R = 1.

E x a m p l e 2. Let F be a field. Define R ' = F ( x , y ) , R i = F(x)[y), R2 = F(y)[x]i R — R i H R2 = F [ x , y ] , T h e n t h e ring R is t h e pullback of R i a n d R2 over R ' in t h e c o m m u t a t i v e square (1) with inclusions ik a n d jk none of which is surjective. W e claim t h a t in this case our results are not true.

By [2, P r o p o s i t i o n 4.1.5, Corollary 4.3.8], we have w d i ? = lgld i î = 2 a n d wd Rk — lgld Rk = 1. Since these dimensions are finite, we have t h a t

3

(5)

l F F D i t ! = l F P D i ? = 1 F I D R = 2 a n d 1FFD7?* = l F P D i E * = l F I D i îf c = 1.

It is easy t o check t h a t t h e / ï - m o d u l e s Rk are flat. So t h e a s s u m p t i o n s of T h e o r e m s 2 a n d 3 hold for n = 1, b u t their conclusions are false. T h e s a m e observation c a n b e m a d e a b o u t t h e estimâtes [1, P r o p o s i t i o n 5, 6 and 7, T h e o r e m s 9 a n d 10, Corollaries 12 a n d 13], which are not t r u e in this case either.

It can b e explained by t h e fact t h a t t h e surjectivity condition cannot be d r o p p e d in t h e basic resuit [1, T h e o r e m 1]. Indeed, we see a t once t h a t t h e .R-module M — R / ( x R + y R ) is neither projective nor flat, whilst t h e Rk- modules Rk <S>r M = 0 are projective a n d flat. From [2, Proposition 3.2.4]

we conclude t h a t t h e R - m o d u l e X = H o mz( M , Q / Z ) is not injective, t h o u g h t h e i?f c-modules H o mf i( / 4 , X ) ~ Homz(i?f e A f , Q / Z ) = 0 are injective.

A c k n o w l e d g e m e n t . T h e a u t h o r would like t o t h a n k Professor A. I. Generalov for suggesting t h e problem and guidance.

R e f e r e n c e s

[1] N . KOSMATOV. On the global a n d weak dimensions of pullbacks of non-commutative rings. Publications M a t h é m a t i q u e s de l ' U F R Sciences et Techniques de Besançon, Théorie des Nombres, Années 1 9 9 6 / 9 7 - 1997/98, 1 - 7 (Univ. de Franche-Comté, Besançon, 1999).

[2] C . A . W e i b e l . An introduction to homological algebra. Cambridge, C a m b r i d g e Univ. Press, 1994.

E q u i p e de M a t h é m a t i q u e s Université de F r a n c h e - C o m t é 16, r o u t e de G r a y

25030 Besançon Cedex France

D e p a r t m e n t of M a t h e m a t i c s a n d Mechanics S a i n t - P e t e r s b u r g S t a t e University

Bibliotechnaya pl. 2

Saint-Petersburg, 198904, Russia E-Mail: koko@nkl442.spb.edu

Références

Documents relatifs

The claim of the second part of the theorem can be reformulated as fol- lows. Derived categories of orthogonal admissible linear sections of homologically projectively dual

From the definition of the Cohen-Macaulay homological dimensions, the connection to Cohen-Macaulayness is less than obvious, but as described in the introduction, we shall prove

The main techniques in [15] are utilizing Hecke algebra structure developed in [13, 14] and simultaneously applying left and right Bernstein–Zelevinsky derivatives, based on

The SYZ proposal [40] suggests that the mirror should be given by the moduli space of Lagrangian torus fibers equipped with flat U (1)-connections.. This is usually called the

In this paper, we mainly consider serval homological dimensions of crossed products such as global dimensions, finitistic dimensions, and strong global dimensions (defined in

(2) use these -filtered modules, which play almost the same role as projective modules for homological calculations, to obtain upper bounds for projective dimensions and

§4 studies the canonical model for the canonically bounded fil- tered (chain) complex, an idempotent Hom-finite theory, and introduces.. its spectral sequence ; some

compactification. [Bestvina et al., 1986] present examples of topologically complete separable ANR’s that satisfy the discrete carriers property and the discrete 2-cells