Bourbaki

Proposition 1

Tag 02HT, permanent. Cite it as [Bourbaki, Tag 02HT].

Théories spectrales, chapter II, § 1. Transformation de Fourier

The canonical mapping of G^\widehat{G} into X(L1(G))\mathsf{X}(L^1(G)) is a homeomorphism.

Let us denote this canonical mapping by ev and, for χ∈G^\chi \in \widehat{G}, let us denote by evχ_{\chi} the image of the character χ\chi under ev, that is to say, the hermitian character f↦χ(f)f\mapsto \chi (f) of L1(G)L^1(G). Considered as a mapping of G^\widehat{G} into the dual of L1(G)L^1(G), the mapping ev is the composite of the injection of G^\widehat{G} into L∞(G)L^{\infty}(G), endowed with the weak topology σ(L∞(G),L1(G))\sigma (L^{\infty}(G),L^1(G)), and the injection of L∞(G)L^{\infty}(G) into the dual of L1(G)L^1(G), endowed with the topology of simple convergence. Since G^\widehat{G} is a bounded subset of L∞(G)L^{\infty}(G), the first mapping is continuous by the Lebesgue theorem (INT, IV, §3, Noo7^o7, Th. 6). The second is likewise continuous, by definition. This proves that the mapping ev is continuous.

If χ∈G^\chi \in \widehat{G} and if f∈L1(G)f\in L^1(G), we have evχ(εx∗f)=χ(x_{\chi}(\varepsilon_x*f) =\chi (x)evχ(f)_{\chi}(f). Taking ff such that evχ(f)≠0_{\chi}(f)\not= 0, we deduce that the mapping ev is injective.

Let ζ∈X(L1(G))\zeta \in \mathsf{X}(L^1(G)) and let f∈L1(G)f\in L^1(G) be such that ζ(f)≠0\zeta (f)\not= 0. Define a mapping χ:G→C\chi : G\rightarrow \mathbf{C} by putting, for x∈Gx\in G:

χ(x)=ζ(εx∗f)ζ(f)(3) \chi (x) =\frac{\zeta(\varepsilon_x*f)}{\zeta(f)} \tag{3}

We have χ(e)=1\chi (e) = 1. Since the mapping x↦εx∗f=γ(x)(f)x\mapsto \varepsilon_x*f=\boldsymbol{\gamma }(x)(f) of G into L1(G)L^1(G) is continuous (INT, VIII, §2, no5^o5, Proposition 8), the mapping χ\chi is continuous. It is bounded because, for every x∈Gx\in G, we have

∣χ(x)∣⩽∥εx∗f∥∣ζ(f)∣=∥f∥∣ζ(f)∣ |\chi (x)|\leqslant \frac{\|\varepsilon_x*f\|}{|\zeta(f)|}=\frac{\|f\|}{|\zeta(f)|}

(Theorem 1 of I, p. 29 and INT, VIII, loc. cit.).

Now let B\mathfrak{B} be a basis of the filter of neighbourhoods of ee consisting of compact neighbourhoods. For every V∈BV\in \mathfrak{B}, let gVg_V be a positive continuous function, zero outside V and of integral equal to 1 (Lemma 1 of II, p. 200). For every function h∈L1(G)h\in L^1(G), one then has

εx∗h=\varepsilon_x*h= limV,B(εx∗h)∗gV=_{V,\mathfrak{B}}(\varepsilon_x*h)*g_V= limV,B(εx∗gV∗h)_{V,\mathfrak{B}}(\varepsilon_x*g_V*h),

in L1(G)L^1(G), the limit being taken with respect to the filter of sections of B\mathfrak{B} (INT, VIII, §4, no7^o7, Proposition 20). In particular, since ζ(εx∗gV∗f)=\zeta (\varepsilon_x*g_V*f) = ζ(εx∗gV)ζ(f)\zeta (\varepsilon_x*g_V)\zeta (f), it follows that

χ(x)=\chi (x) = limV,Bζ(εx∗gV)_{V,\mathfrak{B}}\zeta (\varepsilon_x*g_V).

For every h∈L1(G)h\in L^1(G), we obtain

ζ(εx∗h)=\zeta (\varepsilon_x*h) = limV,Bζ(εx∗gV∗h)=ζ(h)_{V,\mathfrak{B}}\zeta (\varepsilon_x*g_V*h) =\zeta (h) limV,Bζ(εx∗gV)=χ(x)ζ(h)_{V,\mathfrak{B}}\zeta (\varepsilon_x*g_V) =\chi (x)\zeta (h).

Consequently, for x,y∈Gx, y\in G we have:

χ(xy)=ζ(εx∗εy∗f)ζ(f)=χ(x)ζ(εy∗f)ζ(f)=χ(x)χ(y) \chi (xy) =\frac{\zeta(\varepsilon_x*\varepsilon_y*f)}{\zeta(f)}=\frac{\chi(x)\zeta(\varepsilon_y*f)}{\zeta(f)}=\chi (x)\chi (y)

which proves that χ\chi is a homomorphism of G into C∗\mathbf{C}^*. Since χ\chi is bounded and continuous, it is a unitary character of G. Moreover, if g∈L1(G)g\in L^1(G), we have

g∗f=∫G(εx∗f)g(x)dx g*f=\int_G(\varepsilon_x*f)g(x)dx

in L1(G)L^1(G) (INT, VIII, §1, no5^o5, Proposition 7), whence

ζ(g)ζ(f)=ζ(g∗f)=∫Gζ(εx∗f)g(x)dx \zeta (g)\zeta (f) =\zeta (g*f) =\int_G\zeta (\varepsilon_x*f)g(x)dx

=ζ(f)∫Gχ(x)g(x)dx==\zeta (f)\int_G\chi (x)g(x)dx= evχ(g)ζ(f)_{\chi}(g)\zeta (f) (INT, VI, §1, no1^o1, Proposition 1), which shows that ζ=\zeta = evχ_{\chi}. Consequently, ev is surjective, hence bijective.

Finally, let us show that the reciprocal mapping ev−1^{-1} is continuous. Let ζ∈X(L1(G))\zeta \in \mathsf{X}(L^1(G)). Let f∈L1(G)f\in L^1(G) be a function such that ζ(f)≠0\zeta (f)\not= 0. The set W of ξ∈X(L1(G))\xi \in \mathsf{X}(L^1(G)) such that ξ(f)≠0\xi (f)\not= 0 is an open neighbourhood of ζ\zeta in X(L1(G))\mathsf{X}(L^1(G)). For every ξ∈W\xi \in W, what precedes shows that ev−1(ξ)^{-1}(\xi ) is the character

x↦ξ(εx∗f)ξ(f) x\mapsto \frac{\xi(\varepsilon_x*f)}{\xi(f)}

Let F\mathfrak{F} be a filter on W⊂X(L1(G))W\subset \mathsf{X}(L^1(G)) converging to ζ\zeta. Since the set X(L1(G))\mathsf{X}(L^1(G)) is bounded, hence equicontinuous, in L∞(G)L^{\infty}(G), the uniform structure of simple convergence coincides with the uniform structure of compact convergence (TG, X, p. 16, Theorem 1). Let K be a compact subset of G. The set of the εx∗f\varepsilon_x*f for x∈Kx\in K is compact in L1(G)L^1(G) (INT, VIII, §2, No.o5^o5, Proposition 8). We have therefore

limξ,Fξ(εx∗f)=ζ(εx∗f)_{\xi ,\mathfrak{F}}\xi (\varepsilon_x*f) =\zeta (\varepsilon_x*f)

uniformly for x∈Kx\in K. It follows that

limξ,F_{\xi ,\mathfrak{F}} ev−1(ξ)=^{-1}(\xi ) = ev−1(ζ)^{-1}(\zeta ),

hence ev−1^{-1} is continuous at ζ\zeta. This completes the proof of the proposition.

Henceforth we shall identify a unitary character χ\chi of G with the character f↦∫Gf(x)χ(x)dxf\mapsto \int_Gf(x)\chi (x)dx of L1(G)L^1(G).

Bourbaki sets the proof directly after the statement, and the corpus does not separate them, so what is above is the statement and its proof together.

Cited by

Nothing.

Cites

Nothing.

Machine translation of the English, by gpt-5.4, gpt-5-6-mini, against glossary version 34. Not checked by a person. That model is a cut down version of the one the rest of this language was written by.