Intuition The one core idea
A second-order linear ODE is a machine that takes a function y ( x ) and, using its slope and its bending, checks whether it obeys a rule. The entire theory exists to answer: since scaled copies of solutions add together to make new solutions, how few "building-block" solutions do I need to describe every solution?
This page assumes nothing . If the parent note the parent topic used a symbol without telling you what it looks like, we build it here, in an order where each idea leans only on the ones before it.
We will meet, one at a time:
A function y ( x ) — a curve.
The derivative y ′ — its steepness.
The second derivative y ′′ — how the steepness itself changes (bending).
The prime and Leibniz notation — two spellings of the same thing.
Coefficients p ( x ) , q ( x ) and the forcing g ( x ) .
What linear and homogeneous actually mean as pictures.
Constants c 1 , c 2 and initial conditions .
The determinant ⋅ — the two-line grid.
The integral sign ∫ and the exponential exp (needed for Abel).
Let's earn each one.
Writing y ( x ) means: pick an input number x , and the rule hands you back one output number y . Plot the input left-to-right (the horizontal x -axis) and the output up-down (the vertical y -axis), and you get a curve .
Think of x as time and y as the height of a mass on a spring . As time flows, the mass traces a wiggling curve. That curve is the function.
Intuition Why the topic needs it
The whole chapter is about finding which curves obey a certain rule about their steepness and bending. So the very first object is a curve y ( x ) we can look at.
Look at the curve. At any single point, zoom in until the curve looks like a straight ramp. How steeply does it climb? That number is the derivative.
y ′
y ′ ( x ) (read "y -prime") is the slope of the curve at the point x : how many units y rises for each unit x moves right. Uphill ⇒ y ′ > 0 . Downhill ⇒ y ′ < 0 . Flat top or bottom ⇒ y ′ = 0 .
Intuition Why the topic needs it
A spring pulls back based on how fast/where the mass is. "How fast" is exactly a derivative. Every physical law that mentions rate is secretly a y ′ .
Now do the same trick again, but to the slope . As you walk along the curve, the slope itself changes. How fast is the slope changing? That is the second derivative.
Definition Second derivative
y ′′
y ′′ (read "y -double-prime") is the derivative of the derivative — the rate at which the steepness changes. It measures curvature / bending :
y ′′ > 0 : curve bends upward (smile, cup that holds water).
y ′′ < 0 : curve bends downward (frown, spilling cup).
y ′′ = 0 : momentarily straight.
Intuition Why "second-order"
The order of an ODE is the highest derivative it contains. This chapter's equations reach y ′′ — that's what "second-order" names. Physically y ′′ is acceleration (rate of change of velocity y ′ ), and Newton's F = ma is why springs, circuits, and pendulums all land in this chapter.
You will see the same object written two ways. Neither is "more correct" — they are the same steepness.
Definition Prime and Leibniz notation
Prime
Leibniz
Reads as
y ′
d x d y
rate of change of y per unit x
y ′′
d x 2 d 2 y
rate of change of the slope
The d is a shrunken "difference": d x d y is the "tiny run tiny rise " from the limit in §2. Prime notation just abbreviates it.
d x 2 d 2 y means squaring something."
Why it feels right: it has two little 2's. The fix: the top 2 counts how many times we differentiated (d applied twice), and the bottom 2 comes from ( d x ) squared. Nothing is being squared. It is y ′′ , plain and simple.
The parent's standard form is
y ′′ + p ( x ) y ′ + q ( x ) y = g ( x ) .
What are those letters?
Definition The role of each function
p ( x ) multiplies the slope y ′ — think of it as friction / damping that depends on where you are.
q ( x ) multiplies the position y — think of it as a spring stiffness that depends on where you are.
g ( x ) sits alone on the right — an external push (forcing) applied from outside the system.
All three are known functions of x only . They never depend on y . That is precisely what keeps the equation linear (next section).
Intuition Why write it "monic" (leading
1 )
The natural physics equation is a y ′′ + b y ′ + c y = h . Dividing every term by a gives a leading coefficient of 1 and turns b / a , c / a , h / a into p , q , g . We do this once, up front , so every later formula (Wronskian, Abel) reads with a clean y ′′ and no stray a .
These two adjectives decide whether the whole theory even applies, so define them with care.
y is allowed to enter)
The equation is linear if y , y ′ , y ′′ each appear only to the first power , are never multiplied by each other , and are never wrapped inside functions like sin ( ⋅ ) or ( ⋅ ) 2 .
✅ q ( x ) y — a known function times y (fine).
❌ y y ′ — two unknowns multiplied.
❌ sin ( y ) — unknown inside a nonlinear function.
Intuition The visual test for linearity
Linear means: build the output from the input by only scaling and adding. If doubling the input exactly doubles each term it touches, and inputs add without cross-terms, it's linear. Products like y y ′ break this — doubling y quadruples y y ′ .
Definition Homogeneous (is the right side zero?)
Homogeneous means the forcing is switched off: g ( x ) ≡ 0 . The symbol ≡ means "equal for every x ", not just at one point.
Homogeneous: y ′′ + p y ′ + q y = 0 — the system left alone.
Non-homogeneous: g ( x ) = 0 — something is pushing it from outside.
Mnemonic Homogeneous = "home alone"
No outside push (g = 0 ) means the system is home alone , oscillating on its own.
Definition Arbitrary constants
c 1 , c 2
These are numbers you are free to choose — dials. Every second-order ODE leaves exactly two dials open, because "undoing" a second derivative is like integrating twice, and each integration adds one unknown constant.
Definition Initial conditions
Two facts, usually at one starting point x 0 :
y ( x 0 ) = a ( where you start ) , y ′ ( x 0 ) = b ( how fast you start ) .
These two facts turn the two free dials into two equations, pinning c 1 , c 2 to one curve.
Intuition Two dials, two facts
Position tells the mass where to begin; velocity tells it which way and how fast . Together they select a single trajectory out of the whole family — exactly the two numbers the two constants need.
The Wronskian in the parent note is a determinant . Here is what that vertical-bar grid means, from zero.
2 × 2 determinant
For a square of four numbers,
a c b d = a d − b c .
Multiply the main diagonal (a d ), subtract the other diagonal (b c ).
Intuition The picture: an area / a collapse-detector
Read the columns ( c a ) and ( d b ) as two arrows from the origin. The determinant is the signed area of the parallelogram they span.
Area = 0 : the arrows point in genuinely different directions — independent .
Area = 0 : the arrows lie on one line — one is a scaled copy of the other — dependent .
This is exactly why the Wronskian tests independence: it asks whether two solution-arrows have collapsed onto the same line.
Intuition Why the topic needs it
"Are y 1 and y 2 genuinely different solutions, or is one just a stretched copy?" is a collapse question — and the determinant is the collapse-detector. Feeding ( y 1 , y 1 ′ ) and ( y 2 , y 2 ′ ) into it gives the Wronskian W = y 1 y 2 ′ − y 2 y 1 ′ .
Abel's theorem writes W ( x ) = W ( x 0 ) exp ( − ∫ x 0 x p ( t ) d t ) . Two last symbols.
∫ x 0 x p ( t ) d t
The elongated-S sign ∫ means accumulate / total up . ∫ x 0 x p ( t ) d t is the signed area under the curve p from the start point x 0 up to x — the running total of p .
Definition The exponential
exp ( u ) = e u
e ≈ 2.718 is a fixed number. The function exp ( u ) = e u is the curve that equals its own slope (d u d e u = e u ). Its one crucial property here:
e u > 0 for every real u (it is NEVER zero).
Intuition Why these two make Abel's theorem work
W = W ( x 0 ) × ( something never zero ) . So W can only ever be zero if W ( x 0 ) was zero. That is the parent's "all-or-nothing ": check the Wronskian at one convenient point and you know it everywhere. The never-zero exponential is the whole reason.
function y of x - a curve
derivative y prime - slope
second derivative y double prime - bending
coefficients p q and forcing g
two constants and initial conditions
Wronskian - independence test
Superposition and general theory
Once these boxes feel solid, the parent topic is just assembly. For the machinery downstream, see Characteristic equation — constant coefficient ODEs , Method of undetermined coefficients , Variation of parameters , Abel's theorem , and the deeper backing of Existence and uniqueness theorems for ODEs . The "solutions form a vector space" claim rests on Linear algebra — vector spaces and bases , and the single-derivative warm-up is First-order linear ODEs .
Test yourself — say the answer out loud before revealing.
What does y ( x ) draw? A curve: input x horizontal, output y vertical.
What does y ′ measure, as a picture? The slope (steepness) of the curve at a point.
What does y ′′ measure? Bending/curvature — the rate the slope changes; > 0 smile, < 0 frown.
Why is it called a second-order ODE? The highest derivative present is the second, y ′′ .
Are d x 2 d 2 y and y ′′ the same? Yes — two spellings of the second derivative; nothing is squared.
In y ′′ + p y ′ + q y = g , what is g ? The external forcing (push); if g ≡ 0 the equation is homogeneous.
What makes an equation linear ? y , y ′ , y ′′ appear only to first power, no products of them, no sin y etc.
What does ≡ 0 mean? Equal to zero for every x , not just at one point.
Why exactly two constants c 1 , c 2 ? A second derivative needs "two integrations", each adding one constant.
What two facts fix the constants? Initial conditions y ( x 0 ) = a and y ′ ( x 0 ) = b .
What does a c b d equal? a d − b c .
What does a determinant of 0 mean geometrically? The two column-arrows lie on one line — dependent (parallelogram collapses).
Why is e u special for Abel's theorem? It is never zero, so W is all-zero or never-zero.
What does ∫ x 0 x p d t represent? The running signed area under p from x 0 to x .