Intuition The ONE core idea
Parity is a promise: "the number of switched-ON bits will always be even (or always odd)." If a single bit secretly flips on the way, that promise breaks — and by counting we catch the lie. Everything on this page exists to make that counting cheap and automatic.
Before you can understand a parity generator , you must be fluent in the tiny alphabet it is written in. This page defines every single symbol the parent note throws at you — starting from "what is a bit?" — so that not one squiggle arrives unexplained.
A bit is the smallest piece of information: it can be in exactly one of two states . We write those states as 0 and 1 .
Picture a wall switch. Down = OFF = 0 . Up = ON = 1 . That's the whole idea — there is no "half on".
Intuition Why the topic needs it
Parity is a statement about a group of switches : "how many are ON?" You cannot count ON-switches until you agree that a switch is a thing with exactly two positions. The bit is the atom everything else is built from.
A word is an ordered row of bits. We write it left-to-right, e.g. 101 . The little labels d 0 , d 1 , d 2 , … name the individual bits so we can talk about "bit number 2" without pointing.
The symbol d i reads as "the data bit sitting in position i ". The subscript i is just a house-number:
Intuition Why the topic needs it
When the parent note writes d 0 ⊕ d 1 ⊕ ⋯ ⊕ d n − 1 , the "… " means "and every switch in between". Naming each switch d i lets one short formula describe a row of any length n .
The symbol n just means how many data bits there are . If n = 3 the data bits are d 0 , d 1 , d 2 — note the last index is n − 1 , not n , because we start counting at zero.
Parity cares about exactly one number: the count of 1s in a word. Everything else — the order, the positions — is irrelevant to the promise.
A whole number is even if you can split it into pairs with none left over (0 , 2 , 4 , 6 , … ). It is odd if one is always left over (1 , 3 , 5 , … ).
Intuition Why "even/odd" and not the actual count?
Parity throws away how many 1s there are and keeps only whether that number is even or odd — one single bit of information. That is the cheapest possible summary, and it is exactly enough to notice a single flip: flipping one switch always moves the count by one, which always swaps even↔odd.
Picture a switch and its mirror-twin wired to always be in the opposite position.
Intuition Why the topic needs it
Odd parity is literally the complement of even parity — "make it odd" is the flipped version of "make it even". So the parent's odd-parity formula P odd = d 0 ⊕ … is just the even answer with a bar on top. No bar, no odd parity.
This is the tool the whole topic runs on, so we build it slowly.
Definition XOR ("exclusive OR"), symbol
⊕
XOR takes two bits and outputs 1 when they differ , 0 when they are the same .
a
b
a ⊕ b
in words
0
0
0
same → 0
0
1
1
differ → 1
1
0
1
differ → 1
1
1
0
same → 0
Look at the table a second way: a ⊕ b = 1 exactly when an odd number (i.e. exactly one) of the two inputs is a 1 . That re-reading is the bridge to parity.
Intuition WHY XOR and not AND or OR?
We want to answer the question "is the count of 1s odd?" . Test the candidates on the pair { 1 , 1 } (count = 2, even):
AND gives 1 — but the count is even, wrong tool.
OR gives 1 — same problem.
XOR gives 0 — correct: even count → 0 .
Only XOR reports "odd count" honestly. That is the reason the parent picks XOR and nothing else.
An operator is associative if the grouping of brackets doesn't matter:
( a ⊕ b ) ⊕ c = a ⊕ ( b ⊕ c ) .
So we may drop the brackets entirely and write a ⊕ b ⊕ c , working left to right.
Parity of a long word is computed by XOR-ing all its bits one after another. Associativity is the permission slip that lets us do this in any order and still get the same "odd-count" answer. Without it, d 0 ⊕ d 1 ⊕ … would be ambiguous.
⊙ (or just a ⊕ b )
XNOR is the complement of XOR: it outputs 1 when the inputs are the same . It is "XOR, then invert".
Intuition Why the topic needs it
Odd-parity generation asks for the opposite of the even answer, and "XOR then invert" is exactly that. So whenever the parent says XNOR, mentally read "XOR with a bar over the whole thing".
Now every letter in the parent's formulas has a home:
Symbol
Plain meaning
Picture
d i
the i -th data bit
one switch in the row
n
number of data bits
length of the row
⊕
XOR — "differ?" / "odd count?"
the odd-counter gate
x
complement / NOT
mirror switch
X
d 0 ⊕ ⋯ ⊕ d n − 1
"is the data's 1-count odd?"
P
the appended parity bit
the extra safety switch
E
checker's error signal
alarm: 1 = parity broke
Intuition The story in one breath
X asks "is the data odd?". P is the extra switch we flip to fix the count to the agreed even/odd rule. E is what the receiver computes to see whether the rule still holds. Three letters, one idea: keep the 1-count honest.
Complement bar means flip
Associativity lets us chain
Parity generator and checker
A bit can be in how many states, and what are they? Two states: 0 (OFF) and 1 (ON).
In the word d 2 d 1 d 0 = 1 0 1 , what is d 1 ? d 1 = 0 (the middle bit).
If a word has n data bits, what is the index of the last one? n − 1 (we start counting at 0 ).
What does "even" mean for a count of 1s? It splits into pairs with nothing left over (0 , 2 , 4 , … ).
What is 1 ? 0 — the overbar flips the bit.
When does a ⊕ b equal 1 ? When a and b differ (equivalently, an odd number of them are 1 ).
Why XOR rather than OR to test "odd count"? On { 1 , 1 } OR gives 1 (wrong, count is even) but XOR gives 0 (correct).
What does associativity of XOR let us do? Chain many XORs in any grouping/order and get the same answer.
What is X in the parity formulas? X = d 0 ⊕ ⋯ ⊕ d n − 1 — "is the data's 1-count odd?".
XNOR is XOR followed by what? An inversion (complement) — output is 1 when inputs are the same.
What does the checker's E represent? The error signal computed from all received bits including P .