4 ms·

Diagram it and find out

• 082349872349872 3y ago
  Now you all have nerd sniped me. The question of whether a sentence be grammatical equivales the existence of a parse tree which, when linearised, yields the original sentence. Because the sentence is finite, it has a finite number of components, and hence a finite number of possible parse trees, so we could enumerate them all until we find one that works. Exercise: what is the maximum number of parse trees a sentence of length N might have? However (being an english speaker and using wetware as an oracle) a suitable witness can be found by inspection: and ├─can buy │ ├─Money │ └─of │ ├─if not │ │ ├─if not │ │ │ ├─many │ │ │ └─most │ │ └─all │ └─that │ ├─the things │ └─make │ ├─people │ └─happy └─if/then ├─doesn't [buy] │ ├─it │ └─[them] └─is ├─the fault └─ours QEF
  • doublerabbit 3y ago
    QEF? Don't you mean QED?
    • 082349872349872 3y ago
      Having constructed a specific witness (sufficient for existences) rather than a general argument (necessary for universals), I thought "faciendum" more appropriate than "demonstrandum".
      • doublerabbit 3y ago
        Ah, fair enough. I've herd of QED but never QEF. You learn something new everyday.