Abstract
There are, generally speaking, only a few reasons speaker cease to use a certain word. One is that the word is no longer useful, often when the concept is no longer culturally relevant, and the other - that they knowingly and repeatedly choose not to use it - is somewhat complicated.
When we speak of taboo, we tend to think of avoidance speech and replacement of words for dangerous animals the likes of bear in Germanic. The speakers do not want to catch the attention of a dangerous or malignant entity.
The hunter lexicon invents new words for prey, and these may supersede the words from standard register eventually. Swearwords, a necessary yet frown-upon part of everyday speech, tend to "oust" their (near-)homonyms.
What we tend to forget is the prohibition of direct unrestricted reference to G*d present in many cultures. While even a benevolent deity may be modelled as potentially dangerous, vocational texts may refer to it directly which lowers the prediction strength of this concept.
I argue that reverence, not fear was the original motivation for the semi-regular lexical replacements of words for fire in the Indo-European languages. Indo-Iranian cultures often revere fire, yet the Proto-Indo-European distinction of animate-inanimate fire is not preserved and a cognate of neither is to be found.
The connection with cleansing power of fire has been made before for the "animate" *h1n̥gʷnís, I intend to present the etymology of *péh2wr̥ as "the cleanser", which makes it effectively equal to the former and shows why there is disparity between our model of PIE culture and our model for lexical replacement concerning this particular concept.