The most commonly cited examples of the consonant p changing to f in language evolution are in the Germanic languages (Grimm’s Law). The classic example is usually Latin piscis but Germanic fisk, Fisch, fish etc. Lesser known are ones like  German Fuchs, English fox and Sanskrit puccha "tail" and German Freund; English friend, Old Slavic prietenu "friend."

However, as I said in a previous post,  these changes are not unique to the Germanic subgroup of Indo-European. Lyle Campbell ("Historical Linguistics: An Introduction" p. 133) cites examples in Finno-Ugrian e.g. Finnish puu; Hungarian fo "tree"; Finnish purki; Hungarian forr "snow flury" etc.

We see it in several of the Indo-Iranian languages especially Ossetian e.g. fondz "five" but Persian panj from a Proto-Iranian *pank’a.

From the internet this additional information on Persian and Indo- Iranian.

1)"Further, Iranian changed stops such as p before consonants and r and v to spirants such as f: Sanskrit pra "forth," Avestan fra; Old Persian fra;..."

It seems to me that there is no hard-and- fast rule about p changing to f in Indo Iranian… actually you don’t quite have one even in Germanic (compare English ‘help’ and Icelandic ‘hjálp’ with German ‘Hilf’) ... but sometimes this change does occur within the Indo-Iranian group.

2) "Apart from the extensive borrowing of Arabic words in Modern Persian, the Iranian languages have scarcely been affected by unrelated languages, with the notable exception of Ossetic, which has been strongly influenced by the neighbouring Caucasian languages."

This is particular piece of information that would tend to cast doubt  on the change of Parsi (Parsee) to Farsi in Modern Persian as the result of an Arabic influence. Furthermore, in Arabic borrowings from Indo-European languages -p- seems to always change to -b- e.g. Petros, Paulus > Arabic Butrus, Botros, Boulos; Latin princeps "prince"> Arabic brinz; Latin pastilla"pastery"> Arabic bstilla "pastry with a chicken-filled mixture."  

Still, another example of where I’veread about this change occuring is in Manchurian where Old Manchu -p- becomes -f- in many modern Manchurian words. I’m not an expert on Manchurian however I believe that I was no doubt  reading a reliable source.  ::)

Several Semitic languages have undergone the /p/—-> /f/ change too.  But what is your point, Brian?

Yes, and?
I don’t think anybody doubts that.

But my point is that one cannot call every instance of p > f "Grimm’s Law", and that is the starting point of our little discussion, which is apparantly going around in circles.

Neither can one say that Grimm’s Law apply to other languages than the Germanic ones. Grimm’s Law is sometimes (weirdly enough) called "the First Germanic Sound Shift". The weird part is the word "first", because all in all there is not really a second one. The so-called "Second Germanic Sound Shift" only applies to High German.
Some references: here,
here (which also explains Verner’s Law, whithout which Grimm’s Law is meaningless), here,  here.

The reason why i mentioned the High German Sound Shift is pretty clear:
PGm. *sla:pan > OHG sla:ffan
PGm. *opan > OHG offan

Describing this PGm *p > OHG f change by ‘Grimm’s Law’ would be both confusing and wrong.

2) "Apart from the extensive borrowing of Arabic words in Modern Persian, the Iranian languages have scarcely been affected by unrelated languages, with the notable exception of Ossetic, which has been strongly influenced by the neighbouring Caucasian languages."

Add to this grammatical features from Arabic (a.o. so-called broken plurals), and words from Turkish, Mongolic, ...

This is particular piece of information that would tend to cast doubt  on the change of Parsi (Parsee) to Farsi in Modern Persian as the result of an Arabic influence.

This particular piece of information tells us the contrary (which means that we probably will have to move away from "particular pieces of information"):
Persian pardis > Arabic firdaus > (back loan) Persian firdaus.

Frank

Flaminius,

I didn’t say that the p to f change occured in Semitic languages. The ones I mentioned were Germanic; Indo-Iranian; Finno_Ugrian and Manchurian (part of the Altaic family with Mongolian and Turkic). It also occured in your native language, Japanese.   8)

Frank,

Re: "Yes, and?  

I don’t think anybody doubts that."  

We basically agree on one thing then.  

[quote author=brian_costello link=board=etymology;num=1127550628;start=0#3 date=09/24/05 at 17:49:33][P to f chnage] also occured in your native language, Japanese.

Don’t expect me to tell you everytime that the sound represented by "f" in Japanese is bilabial fricative, whereas those in your other examples are labio-dental.

And again, what is your point?

Flaminius,

You are talking about a sound which is an allophone of the consonant -f- and would still sound like an English f to any English speaker hearing it for the first time . Most English speakers don’t make the distinction between the v and f of Dutch and German either. To English ears they sound the same. This is a phenomenon which Dr. Patricia Kuhl (an eminent linguist based in my hometown of Seattle) calls the "magnet effect." I cited an example in Japanese regarding ‘hara kiri’ from Guy Deutscher’s book "The Unfolding of Language" where he says that the word went through three sound changes over time from para kiri to fara kiri to hara kiri. He is a reputable linguist of world reknown. Why would you want to disagree with him over something like this ? Also I am unable to replicate the Greek letter phi on my keyboard which is used by some linguists to symbolize this Japanese f-sound you are talking about. All told, you have a point but still,  aren’t you being a little nitpicky?  :-[

It does not matter how [ph] and [f] sound similar for English speakers.  Labio-dental and bilabial are not allophones in Japanese since there has been no [f] in that language.

I repeat: no [f] in Japanese

be it the result of "p to f change" or a legacy phoneme.

If you stick to your point that [ph] and [f] are allophones, could I prevail upon you to ask for online materials that point to the existence of [f] in historical Japanese (some foreign words are incorporated into Modern Japanese with their legacy [f]s)?

Nitpickeous Flaminius

Most English speakers don’t make the distinction between the v and f of Dutch and German either.

I’m not in this discussion, but German w sounds EXACTLY like an English v and most German v’s (except the ones in foreign words) sound EXACTLY like English f.  I know the situation is more oscillating for Dutch, though.  Dutch speakers themselves don’t pronounce those two sounds uniformly.

It does not matter how [ph] and [f] sound similar for English speakers.  Labio-dental and bilabial are not allophones in Japanese since there has been no [f] in that language.

I couldn’t agree more, Flam.  Brian, would you accept it if some Asian guy who doesn’t differentiate an L from an R told you they are the same?  I don’t think so.

Brazilian dude