The Central Idea
The statement
\[ \lim_{x\to a}f(x)=L \]means that \(f(x)\) can be forced arbitrarily close to \(L\) by requiring \(x\) to be sufficiently close to \(a\), without requiring \(x=a\).
The formal definition measures these two kinds of closeness with two positive numbers:
- \(\varepsilon\) measures how close \(f(x)\) must be to \(L\);
- \(\delta\) measures how close \(x\) must be to \(a\).
Distance on the Number Line
The distance between two real numbers \(x\) and \(a\) is
\[ |x-a|. \]Therefore,
\[ |x-a|<\delta \]means that \(x\) lies within a distance \(\delta\) of \(a\).
Output Distance
Likewise,
\[ |f(x)-L| \]measures the distance between the function value \(f(x)\) and the proposed limit \(L\).
The inequality
\[ |f(x)-L|<\varepsilon \]means that the output lies within \(\varepsilon\) of \(L\).
The Formal Epsilon-Delta Definition
We say
\[ \boxed{\lim_{x\to a}f(x)=L} \]if, for every
\[ \varepsilon>0, \]there exists a
\[ \delta>0 \]such that whenever
\[ 0<|x-a|<\delta, \]we have
\[ \boxed{|f(x)-L|<\varepsilon}. \]Read the Definition in Words
The definition can be read as:
For every desired output tolerance around \(L\), we can find an input tolerance around \(a\) that guarantees the output stays inside the desired range.
Why \(\varepsilon\) Comes First
The definition begins with
\[ \text{for every }\varepsilon>0. \]This means the output requirement may be arbitrarily strict.
Someone can demand
\[ \varepsilon=1, \qquad 0.1, \qquad 0.001, \]or any other positive tolerance.
We must then produce an appropriate \(\delta\).
Why \(\delta\) Depends on \(\varepsilon\)
A smaller allowed output error usually requires the input to be closer to the target.
Therefore, \(\delta\) is often chosen as a function of \(\varepsilon\).
The logical order is:
\[ \varepsilon \longrightarrow \delta \longrightarrow \text{guaranteed output closeness}. \]The Meaning of \(0<|x-a|\)
The condition
\[ 0<|x-a| \]excludes the point
\[ x=a. \]This reflects the fact that a limit concerns nearby behavior and does not require the function to be defined at the target itself.
Punctured Neighborhoods
The condition
\[ 0<|x-a|<\delta \]describes a punctured neighborhood around \(a\).
It includes all points sufficiently close to \(a\) while excluding \(a\) itself.
Interval Form of the Input Condition
The inequality
\[ |x-a|<\delta \]is equivalent to
\[ a-\deltacorresponds to
\[ (a-\delta,a)\cup(a,a+\delta). \]Interval Form of the Output Condition
The inequality
\[ |f(x)-L|<\varepsilon \]is equivalent to
\[ L-\varepsilonA Simple Linear Example
Prove that
\[ \lim_{x\to2}(3x+1)=7. \]We want to make
\[ |(3x+1)-7| \]smaller than \(\varepsilon\).
Simplify the Output Distance
We have
\[ |(3x+1)-7| = |3x-6| = 3|x-2|. \]Therefore, to guarantee
\[ 3|x-2|<\varepsilon, \]it is enough to require
\[ |x-2|<\frac{\varepsilon}{3}. \]Choose \(\delta\)
Choose
\[ \boxed{\delta=\frac{\varepsilon}{3}}. \]Then
\[ 0<|x-2|<\delta \]implies
\[ |(3x+1)-7| = 3|x-2| < 3\delta = \varepsilon. \]Therefore,
\[ \boxed{\lim_{x\to2}(3x+1)=7}. \]The Proof Has Two Phases
Epsilon-delta proofs are often easiest to understand as two stages.
- Discovery: work backward to determine a useful choice of \(\delta\).
- Verification: start from \(0<|x-a|<\delta\) and prove \(|f(x)-L|<\varepsilon\).
Discovery versus Presentation
During discovery, it is natural to start with
\[ |f(x)-L| \]and manipulate it until it is controlled by
\[ |x-a|. \]In the final proof, however, we usually state the chosen \(\delta\) first and then verify that it works.
Another Linear Example
Prove
\[ \lim_{x\to4}(2x-5)=3. \]Compute
\[ |(2x-5)-3| = |2x-8| = 2|x-4|. \]Therefore, choosing
\[ \boxed{\delta=\frac{\varepsilon}{2}} \]is sufficient.
Constant Functions
Consider
\[ f(x)=c. \]To prove
\[ \lim_{x\to a}c=c, \]observe that
\[ |f(x)-c| = |c-c| = 0. \]Since
\[ 0<\varepsilon \]for every \(\varepsilon>0\), any positive \(\delta\) works.
Quadratic Examples Require More Care
Consider
\[ \lim_{x\to2}x^2=4. \]We have
\[ |x^2-4| = |x-2||x+2|. \]The factor
\[ |x-2| \]is directly controlled by \(\delta\), but
\[ |x+2| \]also needs a bound.
Restrict the Input to Control the Extra Factor
We may require
\[ |x-2|<1. \]Then
\[ 1Finish the Quadratic Bound
Under the restriction
\[ |x-2|<1, \]we have
\[ |x^2-4| = |x-2||x+2| < 5|x-2|. \]To make this less than \(\varepsilon\), it is enough to require
\[ |x-2|<\frac{\varepsilon}{5}. \]Choose the Smaller Restriction
We need both
\[ |x-2|<1 \]and
\[ |x-2|<\frac{\varepsilon}{5}. \]Therefore choose
\[ \boxed{ \delta= \min\left\{ 1,\frac{\varepsilon}{5} \right\} }. \]Verification of the Quadratic Proof
If
\[ 0<|x-2|<\delta, \]then
\[ |x-2|<1 \]and therefore
\[ |x+2|<5. \]Hence
\[ |x^2-4| = |x-2||x+2| < 5|x-2| < 5\delta \le \varepsilon. \]Thus,
\[ \boxed{\lim_{x\to2}x^2=4}. \]Why the Minimum Function Appears
A proof may require several conditions on \(\delta\).
Choosing
\[ \delta=\min\{\delta_1,\delta_2\} \]guarantees both
\[ |x-a|<\delta_1 \]and
\[ |x-a|<\delta_2. \]This is a standard technique in rigorous analysis.
The Choice of \(\delta\) Is Usually Not Unique
A valid epsilon-delta proof does not require the largest possible \(\delta\).
Any positive choice that guarantees the desired implication is acceptable.
Simpler choices are often preferable to optimal ones.
The Logical Structure Matters
The formal statement has the structure
\[ \boxed{ \forall\varepsilon>0, \ \exists\delta>0 \text{ such that } 0<|x-a|<\delta \Rightarrow |f(x)-L|<\varepsilon }. \]The order of
\[ \forall \qquad\text{and}\qquad \exists \]is essential.
Quantifiers
The symbol
\[ \forall \]means “for every.”
The symbol
\[ \exists \]means “there exists.”
Therefore,
\[ \forall\varepsilon>0,\exists\delta>0 \]means: for every positive epsilon, there exists a positive delta.
Why the Quantifier Order Cannot Be Reversed
The statement
\[ \forall\varepsilon>0,\exists\delta>0 \]allows \(\delta\) to depend on \(\varepsilon\).
Reversing the order to
\[ \exists\delta>0,\forall\varepsilon>0 \]would demand one fixed \(\delta\) that works for every possible output tolerance, which is generally far too strong.
Implication Language
The symbol
\[ \Rightarrow \]means “implies.”
In a limit proof,
\[ 0<|x-a|<\delta \Rightarrow |f(x)-L|<\varepsilon \]says that input closeness guarantees output closeness.
The Definition Does Not Say \(x\) Equals \(a\)
The condition
\[ 0<|x-a|<\delta \]explicitly requires
\[ x\neq a. \]This is one of the most important distinctions between a limit and a function value.
The Definition Does Not Require \(f(a)=L\)
A limit can exist even if
\[ f(a) \]is undefined or has a different value.
The formal definition only controls
\[ f(x) \]for nearby inputs with \(x\neq a\).
Formal Language for Continuity
Continuity at \(x=a\) requires
\[ \lim_{x\to a}f(x)=f(a). \]In epsilon-delta language, this means: for every \(\varepsilon>0\), there exists \(\delta>0\) such that
\[ |x-a|<\delta \]implies
\[ |f(x)-f(a)|<\varepsilon. \]Because continuity includes the point itself, the puncture is no longer essential in the same way.
Formal One-Sided Limit Language
For a right-hand limit
\[ \lim_{x\to a^+}f(x)=L, \]we restrict inputs to the right:
\[ 0Formal Left-Hand Limit Language
For
\[ \lim_{x\to a^-}f(x)=L, \]inputs must satisfy
\[ 0Two-Sided Limits from One-Sided Limits
Formally, the two-sided limit
\[ \lim_{x\to a}f(x)=L \]holds if and only if
\[ \lim_{x\to a^-}f(x)=L \]and
\[ \lim_{x\to a^+}f(x)=L. \]Formal Language for Infinite Limits
To say
\[ \lim_{x\to a}f(x)=+\infty \]means that \(f(x)\) can be made larger than any chosen positive bound by taking \(x\) sufficiently close to \(a\).
More formally, for every
\[ M>0, \]there exists
\[ \delta>0 \]such that
\[ 0<|x-a|<\delta \]implies
\[ \boxed{f(x)>M}. \]Formal Language for Negative Infinite Limits
To say
\[ \lim_{x\to a}f(x)=-\infty \]means that for every
\[ M>0, \]there exists
\[ \delta>0 \]such that
\[ 0<|x-a|<\delta \]implies
\[ \boxed{f(x)<-M}. \]Formal Language for Limits at Infinity
To say
\[ \lim_{x\to+\infty}f(x)=L \]means that outputs can be made arbitrarily close to \(L\) by taking \(x\) sufficiently large.
Formally, for every
\[ \varepsilon>0, \]there exists
\[ N>0 \]such that
\[ x>N \]implies
\[ \boxed{|f(x)-L|<\varepsilon}. \]Limits as \(x\to-\infty\)
To say
\[ \lim_{x\to-\infty}f(x)=L \]means that for every \(\varepsilon>0\), there exists \(N>0\) such that
\[ x<-N \]implies
\[ |f(x)-L|<\varepsilon. \]A Formal Example at Infinity
Consider
\[ \lim_{x\to+\infty}\frac1x=0. \]We want
\[ \left|\frac1x-0\right| = \frac1x < \varepsilon. \]This holds whenever
\[ x>\frac1\varepsilon. \]Therefore, choose
\[ \boxed{N=\frac1\varepsilon}. \]Formal Proof versus Computation
A formal proof is different from simply computing a limit.
Computation asks:
What is the limit?
A formal proof asks:
Why does the function satisfy the exact definition of that limit?
Formal Proofs Validate the Informal Picture
Tables and graphs suggest behavior.
Algebra often computes the candidate limit.
Epsilon-delta reasoning verifies that the candidate satisfies the rigorous definition.
A Reliable Procedure for Linear Epsilon-Delta Proofs
- Start with \(|f(x)-L|\).
- Simplify until it becomes a constant multiple of \(|x-a|\).
- Solve the inequality for \(|x-a|\).
- Choose \(\delta\) in terms of \(\varepsilon\).
- Begin the formal proof with an arbitrary \(\varepsilon>0\).
- State the chosen \(\delta\).
- Assume \(0<|x-a|<\delta\).
- Verify that \(|f(x)-L|<\varepsilon\).
A Reliable Procedure for Nonlinear Proofs
- Factor or otherwise rewrite \(|f(x)-L|\).
- Identify any extra factors not directly controlled by \(|x-a|\).
- Impose a simple preliminary restriction such as \(|x-a|<1\).
- Use that restriction to bound the extra factors.
- Determine the epsilon-dependent restriction.
- Choose \(\delta\) as the minimum of all required bounds.
- Verify the chain of inequalities carefully.
Example: Full Linear Proof Workflow
Prove
\[ \lim_{x\to3}(4x-2)=10. \]Let \(\varepsilon>0\).
Choose
\[ \delta=\frac{\varepsilon}{4}. \]If
\[ 0<|x-3|<\delta, \]then
\[ |(4x-2)-10| = |4x-12| = 4|x-3| < 4\delta = \varepsilon. \]Therefore,
\[ \boxed{\lim_{x\to3}(4x-2)=10}. \]Example: Full Quadratic Proof Workflow
Prove
\[ \lim_{x\to1}x^2=1. \]We have
\[ |x^2-1| = |x-1||x+1|. \]Require
\[ |x-1|<1. \]Then
\[ 0Therefore,
\[ |x^2-1| < 3|x-1|. \]Finish the Quadratic Proof
Choose
\[ \boxed{ \delta= \min\left\{ 1,\frac{\varepsilon}{3} \right\} }. \]If
\[ 0<|x-1|<\delta, \]then
\[ |x^2-1| = |x-1||x+1| < 3|x-1| < 3\delta \le \varepsilon. \]Therefore,
\[ \boxed{\lim_{x\to1}x^2=1}. \]Common Error: Choosing \(\varepsilon\) Yourself
In a formal proof, \(\varepsilon\) is arbitrary.
You do not get to choose a convenient value such as
\[ \varepsilon=0.1. \]The proof must work for every positive \(\varepsilon\).
Common Error: Letting \(\delta\) Depend on \(x\)
The choice of \(\delta\) may depend on \(\varepsilon\), \(a\), and fixed features of the function.
It should not depend on the particular nearby \(x\) chosen after the \(\delta\)-neighborhood is established.
Common Error: Forgetting the Puncture
For an ordinary limit,
\[ 0<|x-a|<\delta \]excludes \(x=a\).
Omitting the initial \(0<\) can blur the distinction between limits and continuity.
Common Error: Reversing the Quantifiers
The correct order is
\[ \forall\varepsilon>0,\exists\delta>0. \]The demanded output tolerance comes first; the input tolerance is chosen in response.
Common Error: Showing Only a Candidate \(\delta\)
Finding a plausible formula for \(\delta\) is only the discovery stage.
A proof must also verify that the choice actually implies
\[ |f(x)-L|<\varepsilon. \]Common Error: Treating the Definition as a Formula to Memorize
The epsilon-delta definition is a logical guarantee, not merely a string of symbols.
Its meaning is: every desired output precision can be achieved through sufficient input precision.
Practice 1: Interpret the Symbols
Explain in words what
\[ |x-a|<\delta \]means.
Practice 2: Output Tolerance
Explain in words what
\[ |f(x)-L|<\varepsilon \]means.
Practice 3: Punctured Neighborhood
Rewrite
\[ 0<|x-4|<0.2 \]as a union of intervals.
Practice 4: Linear Delta Choice
For
\[ f(x)=5x+1, \]determine a suitable \(\delta\) in terms of \(\varepsilon\) for proving
\[ \lim_{x\to2}f(x)=11. \]Practice 5: Another Linear Delta Choice
Determine a suitable \(\delta\) for proving
\[ \lim_{x\to3}(2x+4)=10. \]Practice 6: Why Use a Minimum?
Explain why a proof might choose
\[ \delta= \min\left\{ 1,\frac{\varepsilon}{6} \right\}. \]Practice 7: Formal One-Sided Condition
Write an appropriate input condition for a right-hand limit at \(x=a\).
Practice 8: Limit at Infinity
For
\[ \lim_{x\to+\infty}\frac{1}{x}=0, \]what value of \(N\) works in terms of \(\varepsilon\)?
Practice 9: Identify the Logical Order
Which comes first in the formal limit definition: choosing \(\varepsilon\) or choosing \(\delta\)?
Practice 10: State the Definition
State the epsilon-delta definition of
\[ \lim_{x\to a}f(x)=L. \]Practice Answers
Practice 1
The input \(x\) lies within a distance \(\delta\) of \(a\).
Practice 2
The output \(f(x)\) lies within a distance \(\varepsilon\) of \(L\).
Practice 3
\[ 0<|x-4|<0.2 \]means
\[ \boxed{(3.8,4)\cup(4,4.2)}. \]Practice 4
We have
\[ |(5x+1)-11| = 5|x-2|. \]Therefore choose
\[ \boxed{\delta=\frac{\varepsilon}{5}}. \]Practice 5
\[ |(2x+4)-10| = 2|x-3|. \]Therefore choose
\[ \boxed{\delta=\frac{\varepsilon}{2}}. \]Practice 6
The choice guarantees both
\[ |x-a|<1 \]and
\[ |x-a|<\frac{\varepsilon}{6}. \]The first may control an extra factor, while the second controls the final epsilon requirement.
Practice 7
A standard right-hand condition is
\[ \boxed{0We need
\[ \frac1x<\varepsilon, \]which is guaranteed by
\[ x>\frac1\varepsilon. \]Therefore choose
\[ \boxed{N=\frac1\varepsilon}. \]Practice 9
\(\varepsilon\) comes first.
The formal order is
\[ \boxed{ \forall\varepsilon>0,\exists\delta>0 }. \]Practice 10
We say
\[ \lim_{x\to a}f(x)=L \]if, for every \(\varepsilon>0\), there exists a \(\delta>0\) such that whenever
\[ 0<|x-a|<\delta, \]we have
\[ \boxed{|f(x)-L|<\varepsilon}. \]What to Remember
The formal definition of
\[ \lim_{x\to a}f(x)=L \]is
\[ \boxed{ \forall\varepsilon>0, \ \exists\delta>0 \text{ such that } 0<|x-a|<\delta \Rightarrow |f(x)-L|<\varepsilon }. \]- \(\varepsilon\) measures output tolerance.
- \(\delta\) measures input tolerance.
- The target input itself is excluded in an ordinary limit.
- \(\delta\) may depend on \(\varepsilon\).
- The order of the quantifiers matters.
- Linear proofs usually give \(\delta\) directly.
- Nonlinear proofs often require auxiliary bounds and a minimum choice.
- Formal definitions extend naturally to one-sided, infinite, and infinite-input limits.
The central meaning is: every desired level of output accuracy can be guaranteed by taking the input sufficiently close to the target.
Where We Go Next
Formal limit language gives us a rigorous way to describe approach and continuity. The next step is to connect these ideas to real systems, where continuity often represents stability, persistence, and predictable response to small changes.
The next lesson, Continuity in Applications, examines how continuity appears in models, measurements, physical systems, engineering, economics, and other applied settings.