The "word replacement strategy" in contextual analysis

[Sekha]

I have seen in this sub-forum a particular strategy in testing a possible meaning of a word by contextual analysis. It consists, having first decided what the word must mean, in replacing it by the chosen term in all its occurrences and checking if the result of the operation gives an intelligible sentence, in which case the term is considered as being the correct meaning.

This strategy is erroneous, and it is dangerous. Here is the demonstration:

In every sentence where the word "dog" occurs, I can replace it by "animal". Sometimes, it will result in an intelligible sentence which forms a true statement:

All dogs have a physical body. >> All animals have a physical body.

But sometimes, although the sentence is intelligible, the statement is completely erroneous:

All dogs have four legs. >> All animals have four legs.