I know that the consonant d undergoes devoicing at the end of a word or a morpheme. This makes it sound exactly like /t/. But what I would like to know from you folks especially are the general rules of pronouncing d. In all other cases, is it pronounced like /d/, the same voiced consonant found in English? Or are there other exceptions outside of terminal devoicing where d is pronounced differently?
You can't talk about the pronunciation of »d« without talking about »t« too.
Rule of thumb:
- In German the letter »d« is pronounced as [d] (voiced alveolar stop)
like in these English words:
- down [daʊn]
- dash [dæʃ]
- sand [sænd]
- lid [lɪd]
- In German the letter »t« is pronounced as [tʰ] (aspirated voiceless alveolar stop)
like in these English words:
- time [tʰaɪm]
- tick [tʰɪk]
- fat [fætʰ]
Some brief information about [tʰ]:
- Do not mix up the aspirated t [tʰ] (like in tee [tʰiː]) with the two english th-sounds [ð] (like in "this" [ðɪs] and [θ] like in thief [θiːf]. The small superscripted h stands for a short breathing sound after the stop-sound. This is called "aspiration".
- In German and in English the sounds [p], [t] and [k] normally are aspirated ([pʰ], [tʰ] and [kʰ]), but many dictionaries (among them wiktionary) doesn't show this aspiration. [p], [t] and [k] are not aspirated in Italian for example, which gives [p], [t] and [k] a softer sound in Italian words, but still different from their voiced siblings [b], [d] and [g].
But there is more to say:
This is a mechanism in many, but not all languages. Turkish has this mechanism, English doesn't have it. In German it depends on the region. In most parts of Germany (north of Bavaria) you can hear this terminal devoicing. In Austria you won't hear it.
This mechanism turns [b], [d] and [g] into [p], [t] and [k] if it would be the last consonant of a syllable.
So in Hannover, Hamburg and Berlin (all three in the northern parts of Germany) »d« at the end of a syllable will be pronounced as if it was a »t« (but I think without aspiration, I'm not sure about this point):
- Fahrrad [ˈfaːɐ̯ˌʀaːt]
- Schwimmbad [ˈʃvɪmbaːt]
- Südfrucht [ˈzyːtˌfʀʊχtʰ]
But you don't hear this final-obstruent devoicing in the south. In Bozen (in the north of Italy), Graz and Vienna (both in Austria) you hear:
- Fahrrad [ˈfaːɐ̯ˌʀaːd]
- Schwimmbad [ˈʃvɪmbaːd]
- Südfrucht [ˈzyːdˌfʀʊχtʰ]
There even is another regional difference in the pronunciation of »d« and »t«:
In the south, the [tʰ]-sound is much softer and even less aspirated than in the north. It is very close to [d]. This is true for standard pronunciation, but the effect is even stronger in dialects.
Take the word »Teppich« (carpet) as an example. If you look for the pronunciation in a dictionary, you will find the northern standard, which is:
[ˈtɛpɪç] or [ˈtʰɛpɪç]
This means: a hard and voiceless t, a short e, and a hard and voiceless p
If you listen to people in Vienna, speaking their local dialect, you will hear:
a soft voiced d, a long e, and a soft voiced b
And this local dialects have a strong influence to pronunciation in standard German. So in Austria you will find, that also t often is pronounced like d. In Austria there is not much difference between Dorf (village) and Torf (peat, turf) or between Mandel (almond) and Mantel (coat).
The fact, that t and d in Austria almost (not exactly!) sound equal, makes it harder for Austrian kids to learn the correct spelling of words, and therefore they learn the names »weiches D« (soft D) and »hartes T« (hard T) for the letters. If in Austria an officer asks you to spell your name, and your name contains d or t, then the officer might ask you
Mit weichem oder hartem D?
(With soft or hard D?)
Whereas it is unclear if the letter at the end of the Question should be written as »D« or »T«, because specially in this question, they really sound exactly equal.
This part is off-topic, i.e. not part of the "official" answer, but is is an interesting episode showing how similar T and D are in Austria:
Adolf Hitler was born in Braunau, in Upper Austria (one of the nine states of Austria), near to the border to Germany, because at the time when he was born, his father, Alois Hitler, was a customs officer, working at the Austrian-German checkpoint. Alois Hitler and his wife Klara both was born near Gmünd and Zwettl in the northern parts of Lower Austria (another Austrian state). Alois was born as Alois Schicklgruber, because his mother (Anna Maria Schicklgruber, Adolf Hitlers Grandmother) was not married when she gave birth to Alois. She later married a man, who was thought to be the father of Alois (which still is matter of debates, because also a jewish salesmen from Graz named »Frankenberger« could have been the father of Alois and therefore Grandfather of Adolf Hitler).
But the name of this man was not Hitler [ˈhɪtlɐ]. His name was Johann Georg Hiedler [ˈhɪːdlɐ]. Pronounced with a long i, and a soft and voiced [d].
Johann Hiedler never adopted his wives son Alois Schicklgruber. But when Johann died, there was no legal heir. To make Alois Schicklgruber the legal son of Johann Hiedler, he had to change his name. But because in Austria Hiedler and Hitler almost sound equal, the registrar did not write Hiedler into the documents, but Hitler.
And so Alois Schicklgruber was renamed to Alois Hitler, and so the name of his son is neither Adolf Schicklgruber, nor Adolf Hiedler, but Adolf Hitler.
There is no standard dialect as you assume. German pronounciation heavily depends on the region you are in, and on the speaker — because people are taking their home dialect to the place where they work and live later on. To give two extreme examples on the d:
In Säggsisch (Sächsisch; Saxon), there's one important pronouciation rule: ''Die Weijen besiejen die Haaden.'' (Die Weichen besiegen die Harten; The soft defeat the stiff.) Nearly all consonants are voiced, or simply left out. As Saxons had a great share in the GDR population and were kept from mixing with other dialects during the iron curtain times, that dialect is sticking out stiffly nowadays, which is a bit paradox.
On the other hand, in Hamburgisch (Hamburg area high german), the Auslautverhärtung is taken to an all-time high by doubling the stiffened t at the end. So Rad becomes Ratt. Similar for other consonants. That one sticks out because former chancellor Schmidt talked that way very purposely to account for his hanseatic origin.
German /d/ is normally voiceless. Voice is an optional feature that commonly occurs in Northern German if the /d/ is inbetween other voiced sounds (vowels, for instance). In Southern German, the optional voicing of /d/ is less common.
The most perceptible feature between the optionally voiced /d/ and /t/ in Northern German is that the latter is aspirated. The voicedness is not the feature that distinguishes the two sounds. Both may be voiceless, but only /t/ may be aspirated. This is very much like in English.
In Southern German, both /d/ and /t/ are voiceless and unaspirated. The distinction is traditionally said to rely on articulary force (hence the terms fortis and lenis). Another more easily measurable explanation is that the distinction relies on duration: /t/ is longer than /d/ (in the traditional explanation, this is an effect of the greater force). Even though the two sounds are distinct, they are difficult to tell apart because the distinction disappears when they are adjacent to another voiceless sound. When speaking standard German, people from the South often use the Northern German aspiration nowadays (less often in Austria).
In traditonal Central German, /d/ and /t/ have completely merged. There are areas where /d~t/ sound is tapped between vowels (as in American English).