Key examples
Key-on-keyring (pattern-matching):
You have 10 keys. You must try each one sequentially. Pick key 1, insert, turn—doesn’t fit. Remove. Pick key 2, insert, turn—doesn’t fit. Remove. Continue until key 7 finally works. This exemplifies true pattern-matching: serial search where each attempt must complete before the next begins.
Cat-seeing-mouse (pattern-recognition):
A cat sees movement in peripheral vision. Instantly, simultaneously: pupils dilate, body lowers, muscles tense, attention focuses, hunting-patterns activate. No sequential checking: “Is it prey? Let me compare to memory. Yes, it matches. Now activate hunting response.” The mouse-pattern and hunting-response are unified—the recognition IS the response.
Traditional computer vision (pattern-matching):
Extract features (edges, shapes, textures) → Compare to stored templates → Calculate match scores → Select best match → Output classification. Each step waits for the previous. Information flows through a pipeline.
Frog’s eye detecting flies (pattern-recognition):
Lettvin et al.’s famous 1959 study: “What the frog’s eye tells the frog’s brain.” The frog’s retinal ganglion cells perform sophisticated pattern-recognition directly. Small dark object moving against light background = immediate tongue-strike. No central processing of “is this food?” The pattern detection and response are unified at the neural level.
Expert systems (pattern-matching):
IF patient_temperature > 38.5 THEN
CHECK infection_markers
IF markers_elevated THEN
RECOMMEND antibiotics
Each rule evaluated sequentially. Must check temperature before checking markers.
Expert intuition (pattern-recognition):
Experienced physician sees patient, instantly recognizes sepsis pattern before consciously processing individual symptoms. Face color + breathing pattern + posture + smell = immediate diagnosis. Dreyfus’s analysis: experts don’t compute through rules but respond immediately through embodied skillful coping.
Spam filter (pattern-matching):
Check subject for “Nigerian prince” → if found, add +10 spam score Check sender domain → if not in whitelist, add +5 spam score Check body for “inheritance” → if found, add +8 spam score If total score > 15 → mark as spam
Sequential evaluation.
Musician recognizing melody (pattern-recognition):
First three notes play → instant recognition of entire piece, composer, emotional tone, what comes next, associated memories. No sequential checking against mental database of melodies.
1