The following is my best reconstruction of the 'proper' way to pronounce Hebrew letters and vowels, according to the text and the Masoretic pronounciation notes (niqud). When I pray or read the Torah, this is how I do so. The pronounciation is based on several logical principles and not on historical research, but it turns out (with one exception) to be very close to scholars' reconstruction of the accent used by the Tiberian Baalei Hamesorah. (Note that it is very unlikely that there ever was one single 'correct' pronounciation used by all Jews - see the Book of Judges. However, since we rely on the Baalei Hamesorah already, I think it most correct to use the pronounciation most consistent with their notations and the actual letters.)
The principles used in this construction are as follows:
| Letter | Regular/Dotted | Undotted |
| 'Alef | Glottal stop ' | |
| Beth | Voiced bilabial plosive b | Voiced labiodental fricative v Note: Although the more logical choice would might been the voiced bilabial fricative (as Ancient Greek "beta"), this is not at all attested in Hebrew; all modern pronounciations have /v/ (except Babylonian which has aspirated /b/, but this seems borrowed, since no other Semitic languages use aspiration). |
| Gimmel | Voiced velar plosive g | Voiced velar fricative gh The Yemenite "jimmel" seems a clear borrowing from Arabic, as no other Semitic languages have it, and it breaks the otherwise clear consistency of dagesh hazaq. |
| Daleth | Voiced dental plosive d | Voiced dental fricative dh |
| Heh | Voiceless glottal fricative h | |
| Waw | Labiovelar approximant w This pronounciation is attested among Yemenites, and is the more plausible by the use of waw as a mater lectionis for shuruq and holam. | |
| Zayin | Voiced alveolar fricative z | |
| Heth | Voiceless pharyngeal fricative | |
| Teth | Voiceless retroflex plosive This admittedly is a tricky one - the other contender was voiceless pharyngealized dental plosive, analagous to sade, but I couldn't pronounce it (so I'm not perfect!). (This is quite similar, in any case...). | |
| Yodh | Palatal approximant y | |
| Kaph | Voiceless velar plosive k | Voiceless velar fricative kh |
| Lamedh | Dental lateral approximant l Pronounced more forward in the mouth than English /l/, but pretty close. | |
| Mem | Bilabial nasal m | |
| Nun | Alveolar nasal n | |
| Samekh | Voiceless alveolar fricative s | |
| `Ayin | Voiced pharyngeal approximant | |
| Peh | Voiceless bilabial plosive p | Voiceless labiodental fricative f Although (as for beth) the more logical choice might have been the unvoiced bilabial fricative (as Ancient Greek "phi"), this is not at all attested in Hebrew; all modern pronounciations have /f/ (except Babylonian which has aspirated /p/, but this seems borrowed, since no other Semitic languages use aspiration). |
| Sadi | Voiceless pharyngealized alveolar fricative | |
| Qoph | Voiceless uvular plosive q | |
| Resh | Alveolar flap r A trill is also acceptable. | |
| Shin | Voiceless palato-alveolar fricative sh | Voiceless alveolar fricative s This violates the "one sound, one letter" principle, but what can I do? The close relation of samekh and shin is clear from the book of Judges, in any case. Perhaps sin should be a voiceless retroflex fricative, but this is not attested in any Semitic languages, to my knowledge. |
| Taw | Voiceless dental plosive t | Voiceless dental fricative th |
Note that I do not use a qamas qatan, since the Baalei Mesorah saw fit to use one symbol, not two. Further, there are no "full" vs. "incomplete" versions of vowels caused by the insertion of yod or waw, since it is pretty clear that those letters are simply matres lectiones, inserted to indicate (pretty much) what vowel is there.
To give an idea of what this sounds like in practice, here is me chanting: