How does one extend the sound of the daleth in the echod of the shema as prescribed in OC 61 (6)?. The letter as we say it does not seem to allow extension. See Dalet (Wikipedia) which defines it as a "a voiced alveolar plosive" and Voiced dental and alveolar stops (Wikipedia) which defines the last term.
It is possible to prolong a plosive (stop). You will hear some who are very careful with reading the prayers or the Tora pronounce a dagesh chazak (dagesh forte) as a geminate consonant; this is a longer consonant, and can be done even with a stop. (If you ever hear someone speaking Arabic or Italian (tutto), you can hear geminate stops also. Many other languages have them, too.) Essentially, the voice box keeps trying to operate (in the case of a voiced stop like dalet), but whatever is stopping the airflow (the tongue in the case of dalet) stops it for longer. A book on phonetics will explain it better; I recommend Ladefoged, A Course in Phonetics.
However, I don't know that that's what the halacha is referring to when discussing the dalet of "echad". As always, CYLOR with practical questions.
Chanoch and Ariel K are correct in their answer, but one can answer at greater length and detail.
The letters beged kefet, בגד כפת are distinguished from other Hebrew letters in taking a dagesh kal, a 'weak' dagesh, at the start of words or after a shva nach. The function of this dagesh kal is to distinguish between the plosive and fricative versions of the letters:
- b/v (actually, in Mishnaic Hebrew, bh, the voiced bilabial fricative, rather than the labiodental)
- p/f (actually, ph, again unvoiced bilabial rather than unvoiced labiodental)
(There is also beged kaperet, where there is a distinction between a resh in front or the back.)
Over the years, different groups have dropped certain of these distinctions, or mapped them to the closest phoneme in the language in their vicinity.
The daled in echad appears at the end of a word, and does not have a dagesh kal. Thus, it should be the fricative -- like the /dh/ in 'either', just as we see the Yemenites have it. The gemara speaks of elongating it, which strongly suggests that they made this distinction.
What about us? Well, we could either adopt the /dh/ for the word, for the duration of Shema; or we can ignore the Talmudic instruction, since it is no longer relevant for us.
Similarly, I would not worry about putting a pause between Eisev | BeSadcha and Hakanaf | Petil or Tizkeru | VaAsitem (the last according to the Rif's girsa) in Shema, despite Rava's instruction in Berachot 15b, because Rava was operating with different phonology, as I discuss here: http://parsha.blogspot.com/2005/03/daf-yomi-brachot-15b-feh_16.html
See also my post on this matter, when we reached it in Daf Yomi, where I also discuss the surrounding gemara.
We pronounce it like the "th" in the word "the".
During the time of the Gemara, they pronounced it like 'Th' in the. Nowadays, ashkenazim pronounce it differently, so we can no longer extend its sound.
Here what I learn from a Rav on the same question : It is written that we should extend the dalet of the word e'had so we not confuse it with a reish (that would give a'her and not e'had). Strange assertion as the nekudot are differents anyway and the sounds are asunder. In fact it could be that this law is part of il'hot sofrim. So it is about writing not pronouncing. The top of the dalet must be extended so we can not confuse it with a reish ! So it is nothing about extending the sound of the dalet.
It's kind of hard to describe this in print, but it can be done if you allow the gap between the tongue and the palate to slowly fill with air as you pronounce the letter.
The source of this halakhah seems to come from Berakhot 13b.
It is well known that according to Tiberian and Babylonian pronunciation systems, דֿ was pronounced as a voiced dental fricative [ð]. The rabbis of the Babylonian Talmud would have read in (a form of) the Babylonian system. Since [ð] is a fricative, it can be easily lengthened.
It is of note that the Jews of modern Baghdad pronounced דּ and דֿ identically (as a plosive) except in two words: אחד from the Shema and אדוני. On this, see "Baghdad, Pronunciation Tradition" by Nimrod Shatil in the Encyclopedia of Hebrew Language and Linguistics.
It says יאריך בד' but I always thought that it doesn't necessarily mean at the end. You can have the kavonah for aleph before starting to say the word, the kavonah for ches during the eh and the kavonah for daled during the uh. It even makes more sense that you should have the kavonah in your head before you start saying the letter. Imagine saying baruch atah hashem and then thinking about what it meant.
Rav Dovid Grossman, in his Daf-Yomi shiur on daf 13a says that according to the Shulchan Aruch, the sound of the Dalet is not supposed to be audibly extended, since it would result in either a d-d-d, or dah sound, and this would ruin the word echad by making it sound like a different word. Rather, one should extend their kavanah beyond pronouncing the letter Dalet, by keeping in mind that hashem is one in all the four directions. This extra kavana is considered elongated, but the sound of the Dalet is still short.
Also, one should extend the sound of the Chet to make sure they don't rush the Dalet. This is done by having in mind that hashem is one above and below (in the heavens and earth [7 heavens+ 1 earth = 8]). This is hinted at by the little point above the letter Chet.
א - Hashem is one
ח - in Shamayim and Aretz
ד - in all four directions