The first letter in “Tao” is an approximation of a Chinese sound that does not have an exact English equivalent. The most accurate rendition would be a combination “t” & “d” sound, as in “Tdhow.”
Actually, there is an exact English equivalent to that sound. It is the “d” sound, so the correct say to say “Tao” is exactly like “Dow” in “Dow Jones.” There is no need for the above explanation, which is rather contrived and has no basis in reality.
This misconception was created by a Western author who did not really know Chinese. He presented an uninformed speculation (that there was a sound in Chinese between “t” and “d”) as an authoritative fact. Others picked up on it, assumed it must be correct because of his status as an author, and spread the misinformation to even more people.
So why should it be that “Tao” should be pronounced with the “d” sound? There is no other usage in English similar to that. If the author’s explanation is wrong, then what is the correct explanation?
The key to the puzzle is the Wade-Giles romanization system, because it is what gave us the word “Tao.” A feature of Wade-Giles is the representation of the unaspirated-aspirated stop consonant pairs using apostrophes: p, p’, t, t’, k, k’, ch, ch’. In that scheme, “t” without the apostrophe is supposed to be pronounced as “d”, and that is why “tao” is supposed to sound like “dao.”
Such specialized rules are confusing to the layperson, and has caused widespread mispronunciation of numerous romanized Chinese words. This is one reason why it will be phased out over time, to be replaced by the Pinyin system.
Using the new system, people will gradually transition from “Tao” to “Dao,” from “Tao Te Ching” to “Dao De Jing,” and so on. The benefit of this transition is that these words will finally sound the way they look to native English speakers. When this comes to pass, perhaps we will finally be able to retire false explanations involving some imaginary combination of sounds.