The Moving Target Called Language

01Languages do not “decay”

Alex Gendler’s TED-Ed lesson begins with a useful provocation: languages are not machines that slowly wear out. They are populations of habits. Speakers copy, simplify, exaggerate and reinterpret those habits, while communities decide which innovations catch on.

That makes “correct English” a historical snapshot rather than a permanent natural law. The forms that look polished today are survivors of earlier competition, not untouched originals.

02The family tree is built from evidence

Historical linguists compare related languages for recurring correspondences, not isolated word resemblances. If a sound changes in a regular environment across many words, the pattern can point backward toward a reconstructed ancestor.

The cognate table shows why. English three, German drei and Latin tres are not proof by appearance alone; their relationship becomes informative when many such correspondences line up. Reconstruction is disciplined inference under incomplete evidence.

03Change has a clock

English is a convenient case study because its conventional periods are familiar: Old English after the Anglo-Saxon migrations, Middle English after the Norman Conquest, and Modern English from the early modern period onward. The boundaries are labels for gradients, not dates when everyone woke up speaking differently.

Sound shifts, grammatical erosion, new compounds and borrowed vocabulary overlap. A word can preserve an old spelling while its pronunciation moves on. The timeline is therefore a scaffold for asking which subsystem changed, when, and among whom.

04Contact is an accelerant

Geography can split a language into dialects, but contact can pull varieties together or layer new vocabulary over old grammar. English absorbed substantial Norse and French influence while retaining a Germanic core; the result is not a tidy blend but a system with multiple historical strata.

Contact also creates false friends for the family tree. Borrowed words can make unrelated languages look similar, which is why linguists prefer regular sound correspondences and shared grammatical quirks over a handful of memorable terms.

05Thousands of branches, no single trunk in view

Language-family counts vary because linguists disagree about where a dialect ends and a language begins. Still, catalogues describe thousands of living languages distributed across major families and isolates. The scale alone is a warning against treating one prestige language as the default human condition.

The second chart uses approximate family counts as a map, not a census. Classification is a scientific model: it improves as documentation grows and as researchers test whether similarities reflect common descent, contact or coincidence.

06Why the origin question stays difficult

Speech leaves a poor fossil record. We can study writing, artifacts, anatomy and genetics, but none gives a direct recording of the first language. That is why origin stories are vulnerable to overconfidence: the evidence is real, but the missing links are enormous.

What can be reconstructed more securely is the process. Children learn a variable system, communities regularize some patterns, and social networks spread or block them. Language evolution is cultural transmission operating at human speed.

07Evolution is not a synonym for improvement

Newer forms are not automatically clearer, richer or worse. A language can lose an inflection and gain a periphrastic construction; it can borrow a word and create a new register; it can make one distinction less visible while sharpening another.

The lesson’s lasting gift is humility. Every speaker is both archivist and innovator. The grammar in your mouth is a moving compromise among ancestors, neighbors, institutions and the next generation.