Most dictionaries use a system of transcribing sounds called the International Phonetic Alphabet (or IPA). It eliminates most ambiguities so that—if you know how to use it—you can figure out exactly how a word should be pronounced.
The IPA is also useful if you want to make up a fictional language for your story, as you can start with the unambiguous pronunciation of a word and use that to invent the spelling.
American dictionaries, on the other hand, don’t have a standardized system of transcribing sounds. There every dictionary makes up its own phonetic system, and it’s really hard to find any background information on any of them.
Since writing this article, I’ve heard various conflicting accounts, some of which suggested that most American linguists do, in fact, use the IPA. Certainly those I’ve interacted with use the IPA, so there’s likely some truth to this.
Keep this in mind throughout this article.
Perhaps the most well-known of these is the one used by the American Heritage Dictionary; it appears to be the only one that has its own Wikipedia page. Today I’ll try and assess its effectiveness in both the study and creation of languages.
What is American Heritage Dictionary Notation?
As I said, there’s very little information on how the American Heritage Dictionary notation system came into existence. My best guess is that it was intended to be easy to learn. This is the system they use to teach American children how to pronounce words.* For that purpose it is perhaps not entirely without merit.
*
I’ve since been informed by some that the AHD system is not used to teach American schoolchildren, although this conflicts with the personal accounts of certain Americans I’ve talked to.
In any case, take that bit with a grain of salt.
Its orthography (spelling system) is clearly based on standard American English orthography, and that might make it easier when teaching it at the same time as standard American English spelling. I do have to wonder how a child would cope with the AHD’s emphases on diacritics, though.
As for adults, I think someone who already speaks American English will probably pick up the AHD’s system pretty quickly. If a sound is usually rendered in English as ⟨a⟩, for example, then American Heritage Dictionary will probably render it as a variant of \a\.
This wouldn’t help a linguist distinguish one sound from another when studying a foreign language, but it does make it easier for the average person to learn. The ⟨oo⟩ in “boot” is represented in the American Heritage Dictionary as \o͞o\. This is where the ambiguities begin, as the ⟨oo⟩ in “foot” is written as \o͝o\; at a glance it’s not as easy to distinguish as [uː] and [ʊ] are.
Phonemic Symbols
Let’s look first at the American Heritage Dictionary system’s ability to transcribe human language. This system isn’t built around a chart like the IPA is; there aren’t places or manners of articulation here. Instead it organizes its symbols in alphabetical order. The few elements of the American Heritage Dictionary that are somewhat organized are inconsistent and don’t make much sense.
Like most American Dictionaries, it has a strange quirk where it groups vowels into “long vowels” and “short vowels.” This may have made sense in the old days when English really did discern a short vowel like [æ] from a long one like [æː], but that long vowel developed into [ɛː] a long time ago.
Inconsistent Vowels
The vowels that usually correspond to the letter ⟨a⟩ are almost always written with some variant of \a\, whether it’s the “long” vowel \ā\ (which might actually be a diphthong), the “short” vowel \ă\, or the \ä\ in “father.”
For some reason it also adds a circumflex accent over vowels that precede rhotic sounds. Diphthongs (which are clusters of two vowels) are often represented here with their own symbols, whose use is often idiosyncratic to the point of lunacy.
The common [aɪ] diphthong is written in the AHD as \ī\, simply because it’s often written as ⟨i⟩ in English orthography. Why not just write it as \äē\?
It would be one thing if these odd principles were applied consistently, but this is the American Heritage Dictionary. The diphthong [ɔɪ] is surprisingly rendered as \oi\ when the AHD symbols for [ɔ] and [ɪ] are \ô\ and \ĭ\, meaning it should rightly be \ôĭ\.
Following this trend, we have [aʊ] (the ⟨ou⟩ in “out”), which the American Heritage Dictionary writes as \ou\, despite the fact that in their system it should be something like \äo͝o\.
Back to the ⟨oo⟩ in “boot” and the ⟨oo⟩ in “foot,” there’s also the ⟨u⟩ in “lure,” which they write as \o͝o\ just like the ⟨oo⟩ in “foot.” In IPA, the ⟨oo⟩ in “boot” is [u], the ⟨oo⟩ in “foot” is [ʊ], and the ⟨u⟩ in “lure” is [ʊə] because it’s actually two sounds.
Ambiguous Consonants
It’s when you get to the consonants that the American Heritage Dictionary system becomes even more baffling.
The affricate [t͡ʃ], which is usually spelt ⟨ch⟩ in English, is written in the American Heritage Dictionary as the digraph \ch\, despite its voiced equivalent [d͡ʒ] being rendered \j\. The American Heritage Dictionary doesn’t even have a sound represented by a lone \c\, so why add an “h” at all? It can only make things more confusing.
Interestingly, what I just suggested isn’t too far off from what a system called Americanist Phonetic Notation does, so the American Heritage Dictionary system has absolutely no excuse for this. In case you’re wondering, I’ll be reviewing APN in a later article.
Fricatives and Sonorants
Concerning the American Heritage Dictionary’s system, it only gets worse. Many fricatives are represented by digraphs, and the ways in which they’re differentiated are somehow worse than the vowels.
The fricative [ʃ] is instead written \sh\, forcing the poor American who tries to use this system to bend over backwards to transcribe words like “grasshopper,” which they must render as \grăs′hŏp′ər\.
Transcribing those same sounds in the IPA is as simple as [ˈɡræsˌhɑpə(r)]. It may not look easier to you, but trust me—IPA is way less of a hastle to type.
Likewise, where the IPA uses [n] and [g], the American Heritage Dictionary writes \n\ and \g\. This is all well and good, but the problem is that the sonorant [ŋ], usually spelt ⟨ng⟩, is written as \ng\.
What happens if a word has [ng] in it? What about the [ŋg] in many English dialects? I suppose Liverpudlian is out, then.
But the worst thing about this awful system is how it deals with the dental fricatives ([θ] and [ð] in the IPA). You’d expect them to use \th\ and \dh\ despite the ambiguities, but I guess that just made too much sense for the American Heritage Dictionary.
No—what they actually do is so insane that I had to check over and over to make sure I’d really seen what I had. Here’s what they do: italicization. Yes, really. The voiceless [θ] is rendered as \th\. The voiced [ð] is rendered as \th\. I cannot think of a worse way to differentiate these phonemes.
Sounds in Other Languages
The only thing worse than trying to transcribe American English with the American Heritage Dictionary is trying to transcribe any other language or dialect with it.
Of the innumerable phonemes that aren’t in General American English, the American Heritage Dictionary is capable of transcribing exactly four. Needless to say, it doesn’t do those four any favours. Again, let’s start with vowels.
The IPA vowels [ø] and [œ] are both rendered in American Heritage Dictionary as \œ\. Honestly, the way things have been going, I was a little impressed that the American Heritage Dictionary sprang for the ligature instead of just writing it \oe\ or something. The vowel [y] is written as \ü\, which isn’t too big a problem.
Nasal Vowels
Then we come to the French nasal vowels, which American Heritage Dictionary transcribes in the style of French orthography. To indicate a nasal vowel, the American Heritage Dictionary adds an uppercase “ɴ” after the nasal vowel.
Now, the IPA does similar things, but it always makes it clear that these symbols are diacritics (little accents and the like) and therefore only modify the sound of the preceding symbol. Aspirated sounds, for instance, are rendered in IPA by adding a small “◌ʰ” to the letter like so: [pʰ], [tʰ], [kʰ].
The French word spelt bon would be transcribed as \bôɴ\ in American Heritage Dictionary and [bõ] in IPA. As it’s still a vowel sound, I think the IPA’s method is the better one.
The ⟨ch⟩ in Loch
Lastly we have the voiceless palatal and velar fricatives—no voiced ones because, again, that would make sense. These are the gravelly consonants in Scottish English and German, and they’re quite popular with those who enjoy making up languages.
The German sound in “ich” is [ç] in the IPA, whereas the Scottish sound [x] is the ⟨ch⟩ in “loch.” Both are represented in American Heritage Dictionary with a single digraph: \ᴋʜ\.
The Other Sounds
Every other sound in human language is, without exception, missing from the American Heritage Dictionary. There is absolutely no way this system could hope to transcribe even the simplest sentence in Arabic, and you can forget about any of the !Kung languages!
You wouldn’t even be able to transcribe Irish English using this thing. If the American Heritage Dictionary system were good for anything, it would be American English and nothing else.
Making Conlangs
Trying to transcribe anything with the American Heritage Dictionary’s system is very likely going to be a headache-inducing ordeal. Ambiguities abound, as do unnecessary distinctions.
But what about making up your own language? Well, let me ask you a question? Do you think Tolkien’s languages could have been transcribed with American Heritage Dictionary notation?
Sindarin has phonemes like the Cymraeg ⟨ll⟩ (IPA: [ɬ]), for which Tolkien used the spelling ⟨lh⟩. Khuzdul has aspirated plosives (such as [tʰ] and [kʰ]), various rhotic trills including [ʀ], and possibly even the voiced velar fricative [ɣ], which he spelt ⟨gh⟩. It would have been impossible to create such languages using the system that American dictionaries use.
When you create your language, you’re going to want it to sound a certain way, but using the American Heritage Dictionary will lead to it sounding like American English. American Heritage Dictionary notation seems almost to be built on the assumption that language comes from writing, but in reality it’s the other way round.
I talked in a previous article about how the IPA can be used to make up a phonemic inventory, and you do not want to be limited by the phonemes of English—particularly if it’s only a single dialect. If you don’t want to use the IPA, make up your own system the way Tolkien did; I guarantee whatever you come up with will work better than the American Heritage Dictionary.
Stick to the IPA
While it may have some application in teaching children, that’s where its uses seem to end. For the study of speech sounds, the American Heritage Dictionary is woefully inadequate; it lacks any way of reliably transcribing most of the phonemes in human language.
When trying to find out how to pronounce a word, it has a number of ambiguities that I’m sure would confuse second-language speakers in particular. For creating fictional languages, the American system is horribly limiting. I’m sure there are worse systems out there, but this is the worst I’ve found.
Some will argue that the American system is “more pragmatic,” but I think “barking mad” would be a better description. If you have an interest in linguistics, then the International Phonetic Alphabet is your friend.
There’s a reason I stick with the Oxford English Dictionary, and it’s not just because Tolkien helped write it. It’s also because when I look at a word’s pronunciation, I know exactly how I should pronounce it.