Words of greek origin, wich were writen with ‘theta’, are graphed in English (and other languages) with ‘th’. This is understandable, since the latin alphabet has no single letter to represent the sound of ‘theta’.
But the latin alphabet uses the letter ‘f’ to represent the sound of the greek letter ‘phi’. Why use ‘ph’ instead of ‘f’ to write words of greek origin?

This is just a guess, but perhaps it is an attempt at an etymological spelling. Keeping the “ph” possibly keeps the word close to its origin. In Romanian, for example, the word cinci, which sounds like “cheench” and means five, was at one time spelled quinqui to make it look more like Latin.

In Greek, when you have affixes or conjugations, you can combine letters and sounds to make different letters.

For instance, the Future Indicative of a verb is made with the ending -so (σω). So the Future Indicative of paideuo (παιδευω) is paideuso (παιδευσω). However, the Future Indicative of pempo (πεμπω) is pempso (πεμψω). The pi and the sigma combined to make psi, so psi was treated as a combination of these two letters, as well as a letter in its own right.

Greek does not have a letter for H, but the H sound does exist, and it’s indicated as a rough breathing mark on vowels. For instance, in the verb histemi (ιστημε) meaning to stand, there’s a rough breathing mark on the first iota, so it starts with an H sound. I can’t draw it, but it’s there. histemi is a versatile verb, and it forms a lot of verbs by prefixing a preposition. The preposition apo (απο) gets added to a lot of words as the prefix ap- (απ-). Guess what happens when it gets added to histemi. The resultant verb is aphistemi (αφιστημε - that’s alpha phi iota sigma tau eta mu epsilon). So you see that phi is seen as a combination of pi and the H sound, as well as a separate letter.

While I am no linguist, I dare say that the difference I perceive between “ph” and “f” is that “ph” sounds (a tiny little bit) * thicker *. The modern greek language does not distinguish between “thicker” and “thinner” in any way, but when greeks speak english, the different sound does come through (I think it comes through with every race that speaks english, but I can only speak for myself).

Phi, which is where we get our spellings with “ph”, was originally an aspirated p, just as chi was an aspirated hard c sound (it is not such now, there’s an international phonetic alphabet sign for chi, which is you guessed it… chi! Aspirated hard c sound is represented as a k with a small superscript h to the right ). Theta was similarly an apirated t sound. All of these fricativized into their current sounds (such as ph to the f sound) not until the end of antiquity.

One of the ways we know these letters were originally aspirates is because if you take a look at the equivalents from Melos and Thera, they use Kappa (K)+ Eta (H) where most other places used chi, and Tau (T)+ Eta (H) where other places used theta (early on Eta used to be pronounced with an H sound).

If i recall correctly, most of the words borrowed from Greek came to us through Latin, and at the time Latin borrowed them they had not yet fricativized. Since the Latins would have been transcribing things in their own writing system and not Greek, they heard the pronunciations slightly differently than their unaspirated counterparts. Thus they added an H after T, C, and P to represent Theta, Chi, and Phi. And this is exactly how they wrote theta, chi, and phi – th, ch, ph.

A telling example of this is “pretentious” roman writing to seem “Greek”:

“This situation [hypercorrection by adding h in words where it doesn’t belong] became confused for a while, however, by the arrival into Latin vocabulary of many Greek words that had laryngeal (formed in the larynx) aspirates and aspirated consonants of the kind regularly <b>written with the letters th, ch, and ph</b>; and it seems likely that the linguistic pretentiousness of some Romans who knew the Greek pronunciation actually led them to attempt to pronounce such words in a Greek way, with [h] sounds all over the place, even in cases where the words were not actually Greek at all. In a famous epigram (no. 84) Catullus makes fun of this way of talking, characterizing someone called Arrius as so high-faluting that he said hinsidias for insidias, and <b>chommoda for commoda</b>.”

  • Vulgar Latin, József Herman.

Just to put in context - the word “philosophy” is spelled with “ph” in French and German, but with “f” in Italian and Spanish (among others). Another interesting example is the use of “y” instead of “i”, as in “synonym”, or sinónimo in Spanish.

The OED records various forms of the word “philosophy”:

I don’t know the exact answer, but I suspect the spelling was standardized at some point by lexicographers who wished to reflect the Greek pronounciation. Other languages possibly adopted a spelling that was more consistent with their language, or was already widely used.

Seriously, you hear a difference between the first syllable in phantom and the first syllable in fantasy? If you do, I believe you, but I’ve never heard anyone claim that.

Achernar and Doobieous present consistent explanations for the use of ‘ph’ to represent ‘phi’, by I think the last one is more probable: different pronontiation led to the use of a combination of letters in Latin to represent Greek sounds.
In the contrary, combination of letters in Greek led to the use of single letters in modern words. For instance: Gaia (Earth) and paidos (child) have changed ‘alpha + iota’ for ‘e’. Even in French, where ‘ai’ sounds like ‘e’ we write géographie and pédagogie.

French prefers to go the etymological route. Spanish and Italian tend to go much more phonetic. So, they change ph to f and y to i, although it’s a bit odd that Spanish doesnt just keep the y’s as y since it uses y to represent i (and there is no single word “i” in spanish).

That isn’t quite correct. “Y” is used in Spanish purely as a consonant (I think). When “y” is used as a vowel, it changes to “i.” In Spanish, “y” is called 'i griega," or “Greek i.”

Spanish is pretty strict about spelling things exactly as they sound, with the exception of the silent “h.”

That’s very interesting. It reminds me of English (Cockney?) pronunciation that drops the initial “h” sound in many words. Some people do make the overcorrection in a misguided attempt to sound posh. It’s used and extended wittily in “The Gnu” by Flanders and Swann: “I h’ain’t a h’elk! I’m a g-nu!”

Since Jtull is Greek, it’s perfectly possible to him recognize this differences. I, for instance, although I speak a very good French, cannot perceive the difference between ‘â’ and ‘a’, but a French friend of mine says they sound differently.
By the other way, he can’t perceive the difference between ‘r’ and ‘rr’ in Portuguese, sounds very distinct to any Portuguese speaking person.

I’d say you’re wrong, since “and” - “y” is acting as a vowel, not a consonant. It is not changed to I. THAT is what i was talking about. It does not have a consonantal J /j/ in y, and neither does it become spelled as “i” either. This is what i was saying, if Spanish replaces greek y with i everywhere else, which is obviously an /i/ sound, then why does it not use i (alone) for and, especially since there is no word in Spanish spelt as “i”.
Also, it’s not etymological here from Greek since “y” comes from “et”.

I think that ‘y’ is a semivowel in Spanish. You are wright in saying that ‘y’ alone, meaning ‘and’, is pronounced as a vowel, but before a vowel ‘y’ is pronounced as ‘j’.

Well, it was pretty difficult to tell what you were talking about in the post I quoted. You certainly didn’t seem to be talking about the word “y.”

I will concede there may be some exceptions to “y” always representing a consonant sound in Spanish. But Spanish does always replace the Greek “y,” used as a vowel, with “i.” As you point out, the word “y” is not derived from Greek.

I’m not saying it is much of a difference, but it subtly exists. It may even not be a difference that you can actually hear, but when I look or say an english word like “phantom”, I instictively feel that this “ph” is somehow “thicker” than the “f” in “fantasy”, and that feeling changes the way I treat the word, surely.
I think it’s because we only have one “f” in our language, and since we learn to associate it with the english “f”, seeing “ph” written somewhere makes us try (unconsiously) to differentiate its pronunciation.

On the other hand, we do have a lot of vowels that sound exactly the same nowadays. Five ways to say “i” (iota, omikron+iota, epsilon+iota, ypsilon, hta (H)), two ways for “e” (alpha+iota, epsilon), two different vowels for “o” (omikron, omega).
The sad thing is that the ministry of education has been trying to reduce all these different combinations lately… New school books have been introduced, where many words have all these combinations (where appropriate) substituted with the single vowel. So “traino” (train) has become “treno”, “aygo” (egg) has become “avgo”, etc. This started a long way back in the mid 80s, I can easily remember my elementary school books having a section at the end of each page where they noted the “bad” (old) spelling and the “correct” (new) one of a certain batch of words. Luckily I was one of the people who didn’t absorb the new way. I really enjoy the words with the full set of vowels and I sicken at the sight of the new crippled words.

jtull writes:

> I’m not saying it is much of a difference, but it subtly exists. It
> may even not be a difference that you can actually hear, but
> when I look or say an english word like “phantom”, I instictively
> feel that this “ph” is somehow “thicker” than the “f”
> in “fantasy”, and that feeling changes the way I treat the word,
> surely.

If a distinction in the pronunciation of English doesn’t exist to native speakers of English, it doesn’t exist at all. What you’re doing is convincing yourself that there’s a difference there when there actually is none. That “thicker” and “thinner” distinction doesn’t apply to any sound in English, and you’re imagining things if you think that it does.