Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $g \in L^1[0,1]$. Suppose that given any pair of rationals $0\leq p\lt q \leq 1$, we have $$\int_p^q g(x) d\mu=0.$$ Please I would like help in showing that $g=0$ almost everywhere on $[0,1]$.

share|cite|improve this question
Is $\mu$ just the normal Lebesgue measure on $[0,1]$, or is $\mu$ something else? – JavaMan Nov 28 '11 at 2:38
I presume you want to prove that $g$ is zero $\mu$-a.e. Am I right? – Srivatsan Nov 28 '11 at 2:39
@JavaMan: $\mu$ is the normal Lebesgue measure. – AKM Nov 28 '11 at 2:50
@Srivatsan: Exactly so. Thanks. – AKM Nov 28 '11 at 2:51
Are you familiar with Vitali's covering lemma? – user9352 Nov 28 '11 at 3:00
up vote 5 down vote accepted

Here's one approach.

  • Let $a$ and $b$ be elements of $[0,1]$ with $a<b$. Taking sequences of rationals $(p_n)$ and $(q_n)$ in $[0,1]$ with $\lim\limits_{n\to\infty} p_n =a$ and $\lim\limits_{n\to\infty} q_n = b$, you can show that $\int\limits_a^b g\ d\mu =\lim\limits_{n\to\infty}\ \int\limits_{p_n}^{q_n}g\ d\mu$.
  • If $U=\bigcup\limits_{n=1}^{\infty}(a_n,b_n)$ is an arbitrary open subset of $[0,1]$, with $\{(a_n,b_n)\}$ a pairwise disjoint collection of open intervals, you can show that $\int\limits_U g \ d\mu =\sum\limits_{n=1}^\infty \ \int\limits_{a_n}^{b_n} g\ d\mu$.
  • If $G=\bigcap\limits_{n=1}^\infty U_n$ is an arbitrary $G_\delta$ set, with each $U_n$ an open subset of $[0,1]$ and with $U_{n+1}\subseteq U_n$ for each $n$, you can show that $\int\limits_G g\ d\mu=\lim\limits_{n\to\infty}\ \int\limits_{U_n} g\ d\mu$.
  • If $E$ is an arbitrary measurable subset of $[0,1]$, then $E=G\setminus N$ for some $G_\delta$ set $G$ and some null set $N$, and $\int\limits_E g\ d\mu=\int\limits_G g\ d\mu$.
  • Consider the measurable sets $\{x:g(x)>0\}$ and $\{x:g(x)<0\}$.
share|cite|improve this answer
Would you say you have given him more than mere "help"? – GEdgar Nov 28 '11 at 14:53
Maybe if we defined the "mere help norm", we could answer ^that ;) – The Chaz 2.0 Nov 28 '11 at 17:13

Maybe show the collection of sets $A$ such that $\int_A g\,d\mu = 0$ is a sigma-algebra.

share|cite|improve this answer
Could you please expand a bit. Thanks. – AKM Nov 28 '11 at 14:45
@AKM: Jonas has done your work for you. – GEdgar Nov 28 '11 at 14:51

The following two facts are helpful:

(1) Let $f:[0,1] \rightarrow \overline{\mathbb{R}}$ be an integrable function then for all $\epsilon > 0,$ there exists a $\delta > 0$ such that for all $A \in \mathcal{L}([0,1])$ with $\mu(A) < \delta$ the integral $\displaystyle \int_A |f| d\mu < \epsilon.$

(2) Let $A \in \mathcal{L}([0,1]),$ then for any $\epsilon > 0$ there exists an open set $U \supset A$ such that $\mu(U) - \mu(A) < \epsilon.$ set

Now let's see how these facts can be used to prove your claim. Let $A = \{x\in [0,1]: g(x) \ge 0\}.$ If we can show $\displaystyle \int_A g d\mu < \epsilon$ for all values of $\epsilon > 0$ the claim will follow.

Fix $\epsilon > 0.$ Choose a $\delta > 0$ such that for all $A \in \mathcal{L}([0,1])$ with $\mu(A) < \delta$ the integral $\displaystyle \int_A |g| d\mu < \epsilon.$ Now choose an open set $U \supset A$ satisfying $\mu(U) - \mu(A) < \delta.$

Appealing to the dominated convergence theorem, one observes $\displaystyle \int_U g d\mu$ for any open set $U$ of $[0,1].$

It follows

$$0 = \int_U g d\mu = \int_A g d\mu + \int_{U\setminus A} g d\mu.$$


$$|\int_A g d\mu| \le \int_{U\setminus A} |g| d\mu < \epsilon$$

and the claim follows.

share|cite|improve this answer

Have you tried to use the fact that $\mathbb{Q}\cap [0,1]$ is dense in $[0,1]$?

edit: Actually what I posted above isn't particularly helpful. Here's a different idea -

Let's assume $g\geq0$ on $[0,1]$ for now. Let $A=\{x \,|\, g(x)>0\}$ and lets suppose $\mu(A)>0$. Then we have that $$ \int_A g d\mu > 0$$ but we also have that $A\subset[0,1]$, so $$\int_A gd\mu \leq \int_{[0,1]} gd\mu=0 $$ since $0$ and $1$ are both rational. Perhaps you can extend this to the case where we don't need to assume $g>0$?

share|cite|improve this answer
I'm not sure how to use that fact. Could you please elaborate a bit? Thanks. – AKM Nov 28 '11 at 3:05
@AKM I misread the question the first time around, so I've edited my answer. – Sid Raval Nov 28 '11 at 3:21

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.