Problem pronouncing 'TH'

Matt   Thu Feb 07, 2008 11:36 am GMT
I am British, but still have problems pronouncing 'TH'.

Most notably, I appear to pronounce three as free and I have become very self concious of it.

I have no problem with 'the', 'there', 'these' but words like 'mouth' still vex me. Even my name Matthew, ends up being 'Maffhew'

My mother is French, so not sure whether this has had an impact, or whether I have picked this up from Friends.

People always say they consider my accent to be quite posh, but I can no longer accept my failing with 'th' and it is beginning to bother me.

I practiced last night, yes....I had to practice! and I eventually got it (with much difficulty), but then I lose it again, and in day to day conversation I forget it completely.

Any advice, or views on this?

Thanks :)
Gabriel   Thu Feb 07, 2008 6:22 pm GMT
Isn't this supposed to be a rather common feature of many English accents? It has its own name: "th fronting". Usually it involves the merging of /T/ and /f/ (as in your speech, making "three" sound like "free"), and /D/ and /v/ (so that "brother" sounds like "bruvver"). I've heard it often on British television and movies. So if it is a feature of the dialect around you, I imagine it's not very noticeable, and hardly worth changing.
Guest   Thu Feb 07, 2008 7:06 pm GMT
You shouldn't worry about it. There really isn't anything wrong with that. More than likely most of the people in your region do the same thing.
Guest   Thu Feb 07, 2008 7:11 pm GMT
In America, small children and adults with speech impediments sometimes talk like that.
Travis   Thu Feb 07, 2008 7:26 pm GMT
>>In America, small children and adults with speech impediments sometimes talk like that.<<

Not necessarily. Around here in southeastern Wisconsin, the trend seems to be that little kids preserve /θ/ (unlike Matt) but invariably shift initial /ð/ to [d] and *occasionally* shift medial /ð/ to [v]. Of course, the thing is that such really is prototypical of the pronunciation in the dialect here in general, where it is very frequent for adults to still pronounce initial /ð/ as [d] and one can still hear they sporadically pronouncing medial /ð/ as [v] but /θ/ is largely perserved (except sometimes before /r/, where it may become [t̪]). At least here, pronouncing initial /ð/ as [ð] seems to actually be a later-learned feature that most are not truly consistent with in actual everyday speech (and of whom many express less often than not); I myself remember always perceiving initial [ð] as being "hard" as a little kid, and still only consistently use it with conscious effort on my part.
Damian in Edinburgh   Fri Feb 08, 2008 12:15 am GMT
The use of an "f" or even a "v" for "th" seemed so widespread in London. All the while I was working down there I heard people say fings (sorry - things) like: "I fink you should take a look at vis!" I think it's now accepted part of Londonspak in many people down there! I shouldn't fret too much about it, Matt. I fink you should just go wiv vuh flow.....
D in E   Fri Feb 08, 2008 12:17 am GMT
Lo   Sun Feb 10, 2008 7:53 pm GMT
Well, it if really bothers you as much as you say I guess you can imitate other people's accents. I don't watch the BBC all that often, but the little I've seen, they keep their /D/s and /T/s the way they are. It's not that hard a sound, especially when you already have it for other words. Good luck and make it work!
Guest   Sun Feb 10, 2008 8:42 pm GMT
Isn't the replacement of the two 'th' sounds with 'f' and 'v' very trendy in the UK? Isn't it destined to become the norm over there in a generation or so? Will this simplification eventually spread to AU, NZ, Ireland, CAN, US, etc? If all this is true, you should be proud to be a trendsetter, leading the way to a new form of English.
Damian in Edinburgh   Mon Feb 11, 2008 4:36 pm GMT
***Isn't the replacement of the two 'th' sounds with 'f' and 'v' very trendy in the UK? Isn't it destined to become the norm over there in a generation or so?***

Guest: Short answer - no. It'll never go mainstream - and in any case, I fink (sorry - think) I did a lot of people in the London area (and also those in a far wider area of Southern England in particular) a wee bit of an injustice by suggesting, or at least implying, that they ALL speak that way......very many do NOT. But it's true that a fair number of people do and here again I think it's confined more to a certain section of the population at large, and even amongst them it's largely confined to the more informal conservational trendy it may well be in those circumstances, but in other situations I reckon efforts are made to conform to more "proper" speech.

However, on the council estates of the capital city, say from Lewisham to Lambeth, or Plaistow to Palmers Green, (I'm not specifically targetting those areas - just making a point) you'll find a lot of people who will continue to do the "f" and "v" thing'll always be part of their standard speech, all part of the Estuary "thing" with bits and pieces of ghetto speak chucked in for good measure. That's all part and parcel of any huge metropolitan, extremely diverse and cosmopolitan city, and now spread to suburban and urban areas outside the Greater London area.

It's very unlikely to become a "trend" that will ultimately become universal in the English speaking world.
Travis   Mon Feb 11, 2008 4:51 pm GMT
I doubt that it would at least fully become part of North American English, as the general trend in NAE is to assimilate word-initial /ð/ to sounds in preceding words and, in AAVE and Upper Midwestern dialects, to stop or affricate unassimilated word-initlal /ð/ and, on a more limited basis, /θ/ rather than to front them. Furthemore, widespread realization of medial and final /θ/ and /ð/ as [f] and [v] is generally associated with AAVE here, which would likely be a barrier to the spread of such sociolinguistically, even though I myself have heard medial/final /ð/ realized as [v] in my own dialect at times (but coda /θ/ may be stopped in my dialect at times, on the other hand, and is often stopped in the case of "with").
Travis   Mon Feb 11, 2008 6:52 pm GMT
>>Stopping is much more likely to spread, IMO, since it already occurs in various sociologically neutral varieties.<<

Furthermore, there are very many dialects which do not have stopping proper, but do have gemination of preceding final stops by following initial /D/, which seems to essentially be a "light" version of interdental hardening.
Joe Tun (aka) U Kyaw Tun   Mon Feb 11, 2008 8:54 pm GMT
Problem pronouncing <th>
Dear Matthew
First let me rewrite the digraph <th> as 'thorn' character, which was present in Old English and Middle English. Its symbol is <þ>. It is an ASCII character (press down Alt key and type in 0254) whereas the Greek letter 'theta' is not.
Secondly, it's strange that you have the problem confusing <þ> with <f>. The usual problem is the confusion between <þ> and <s> as is present in mis-pronunciation of the Burmese-Myanmar <þ> as <s> by the Indian and Western transcribers. All the three characters, <f> <þ> and <s> are produced with friction (but without your vocal cords vibrating) and are known as fricatives. If you look at the Places of Articulation (POA) of the three characters, you will find that they are very close: <f> is articulated as labio-dental, <þ> as dental, and <s> as alveolar. They are all articulated very near the lips. <f> is articulated with lower lip touching the upper teeth with the tongue in neutral position. <þ> is articulated with the tip of the tongue lightly held between the upper and lower teeth. Both are articulated without any hissing sounds and are known as thibilants to be differentiated from <s> which is pronounced with a hissing sound. Because of the hissing sound, <s> and the next character represented by digraph <sh>, are known as sibilants.
Thirdly, since your confusion is between <f> and <þ>, it is more understandable than the confusion between <þ> and <s>. Just concentrate on how you articulate the sounds. For <f> the lower lip is very much involved, but for <þ> the lips are quite apart and do not take part in producing the sound.
Fourthly, your name Matthew. You will see the consonant cluster made up of two characters in the middle part of your name. (Remember, I have replaced the second <t> and the following <h> with <þ>. So the consonant cluster is made up of two characters only. The clusters are found in Myanmar and Indic languages and are called ligatures. In Myanmar this particular ligature is written as a vertical ligature, with the <t> over <þ>. In Romabama, the system of transliteration I am developing for the study of phonology of Burmese (Burmese is the spoken language, whereas Myanmar is the script), it is written as <t~þ>. So your name is a disyllabic word, and I will pronounce it in careful speech as: /mæt.þu:/. In the Daniel Jones Pronouncing English Dictionary, 16th ed., 2003, p334 two transcriptions that are different from mine are given.
It is quite rare to find the <þ> in world languages. It is certainly not present in Germanic languages and many Indic languages. I am of the opinion that <þ> in English is the left over of pre-Viking Brittanic languages (I am on very thin ice over this opinion.). If you are interested in linguistics, phonetics, etc., you are invited to go into my website: and look in the section Romabama, or you may correspond directly with me: .
Travis   Tue Feb 12, 2008 3:48 pm GMT
The matter is that different language varieties often tend to eliminate interdentals in different ways. For instance, Cockney and more Cockney-like Estuary have a general shift of /θ/ to /f/ and /ð/ to /v/, AAVE has a shift of medial/final /θ/ to /f/ or /t/ and medial/final /ð/ to /v/ or /d/, a shift of initial /ð/ to /d/, and possibly a realization of initial /θ/ as [t], and Upper Midwestern dialects tend towards shifting /ð/ to [d] and /θ/ to [t] but this greatly varies in practice (as there are dialects which only shift unassimilated initial /ð/ to [d̪] or [d̪ð] while on the other hand there are those which have the full general shift). Outside of English, how languages get rid of interdentals in foreign words likewise varies. For instance, most Germanic languages other than English, Scots, and Icelandic tend towards shifting /ð/ to /d/ and /θ/ to /t/, while on the other hand French tends towards, at least supposedly, shifting /ð/ to /z/ and /θ/ to /s/. There is no one "easier" way to eliminate interdentals in some given language variety, even if some way seems more intuitive to oneself.
Lazar   Tue Feb 12, 2008 4:35 pm GMT
I think Cockney and Cockney-influenced Estuary also have a shift of initial /ð/ to [d].