I feel that Klingon has a bad reputation among conlangers. Even I have mocked it in my previous post for its terrible romanisation system. But is it really that bad? Or is the bad romanisation system making it look worse than it actually is? As a Star Trek fan and as a Conlanger, I feel that I should take a closer look at the Klingon language; and will attempt to give it a fair and balanced review.
Klingon was created by Marc Okrand for the film ‘Star Trek III: The Search for Spock’ in 1984. However, the first use of ‘Klingon’ in Star Trek was actually in the film ‘Star Trek: The Motion Picture’, but this was more gibberish than an actual constructed language. When creating Klingon, Marc Okrand used these lines of gibberish as a basis for a developing a proper conlang. In 1985, a description of Klingon’s grammar and a dictionary was published in ‘The Klingon Dictionary‘ (which I’ll abbreviate to TKD) written by Marc Okrand himself. Later in 1997, Marc Okrand wrote another book called ‘Klingon for the Galactic Traveler‘ (KGT). This included new vocabulary and details about various registers and dialects of Klingon. Those books will be my primary sources for this review.
First I would like to mention the writing style of TKD. It’s immersive, clear and a joy to read, especially the introduction. Considering it’s the first official source describing the Klingon language, it gives a very good first impression.
One problem I have with it is how he uses “this is currently unknown” or “more research needs to be done” to explain away any gaps or oddities. At first, this trick may seem quite clever, but he uses it so frequently that in some places it comes across as a bit lazy.
The KGT is also very well written. Although compared to TKD, it feels a lot more padded out, so I find myself skim reading through sections. It is still a very interesting book though, and I highly recommend it to fellow Star Trek fans.
Unfortunately, TKD does not provide any IPA transcriptions for its phonemes. Instead, phonemes are listed in Klingon’s romanisation system with a description explaining how to articulate them. While I find this completely reasonable considering TKD’s target audience, it would be nice for a more technical description of Klingon’s phonology to be included in something like an appendix. The article on Wikipedia includes an IPA chart, but unfortunately no source is cited for it. So the descriptions I give below are my guesses to what is described in TKD. If what I describe below is not what was intended, then the description in TKD has failed to properly communicate the correct phonetic information. This potential problem could be avoided altogether if IPA transcriptions were provided in some way.
|Nasal||m /m/||n /n/||ng /ŋ/|
|Voiceless Stop||p /p/||t /t/||q /q/||‘ /ʔ/|
|Voiced Stop||b /b/||D /ɖ/|
|Voiceless Affricate||tlh /t͡ɬ/||ch /t͡ʃ/||Q /q͡χ/|
|Voiced Affricate||j /d͡ʒ/|
|Voiceless Fricative||S /ʂ/||H /x/|
|Voiced Fricative||v /v/||gh /ɣ/|
|Approximant||l /l/||y /j/||w /w/|
Wikipedia describes H as being a uvular fricative /χ/, but after reading the descriptions in TKD, I disagree with this. It clearly specifies that gh and H are in the same place of articulation. Also the entry for q states:
To produce Klingon q, the main body of the tongue touches the roof of the mouth at a point further back than it does for gh or H. Indeed the tongue reaches for or touches the uvular.1.1. Consonants – The Klingon Dictionary – by Marc Okrand
This clearly implies that q is a uvular consonant and that both gh and H are velar.
You may have noticed that this consonant inventory is very inconsistent. Phonemes are spread over a large number of places of articulation and there are a lot of gaps. This seems to have been done intentionally in order for it seem less human and simulate an alien vocal tract. However, it looks like consonants have been randomly picked off an IPA chart. A better approach to this would be to describe anatomical differences between the human and Klingon vocal tracts, develop a phonemic inventory based on the alien anatomy, and then try to find the nearest human equivalents.
Klingon has affricates, and a major pitfall with using affricates is asserting them to be phonemic when in fact they should really be analysed as consonant clusters. Fortunately, Klingon doesn’t make this mistake. Firstly, none of the fricative components, /ɬ, ʃ, ʒ, χ/, exist as independent fricatives. Secondly, Klingon’s phonotactics doesn’t allow consonant clusters in the onset. For both of these reasons, it makes perfect sense to analyse these phonemes as affricates.
Doubled consonants can occur in between syllables. Whether these are pronounced as true gemminates or just a spelling convention to show the structure of compounds is not made clear in TKD. However, the KGT clarifies that older generations will pronounce these as gemminates, but younger speakers will sometimes pronounce these as singular consonants. I think this is an interesting way of fixing an oversight of TKD.
I would also like to point out there is no allophonic variants for any of these consonants, they have the same phonetic realisation in every context. I think this is a missed opportunity, since allophony can help make a conlang feel more naturalistic.
|Close||I /ɪ/||u /u/|
To be honest, Klingon’s vowel system is a bit bland. It’s basically just a five vowel system. There are a couple of points of interest though. Firstly, /ɪ/ is used instead of /i/, which I think is a good choice since /i/ doesn’t fit in with Klingon’s intended sound. Also, /o/ seems to only occur in diphthongs. Marc Okrand compares the o to the o in mosaic. Assuming he meant the American English pronunciation, the word mosaic is pronounced /moʊˈzeɪ.ɪk/, which suggests that the o is the diphthong /oʊ/. He might have meant just the onset of that diphthong, but later he states:
No words in Klingon have ow or uw. If they did, they would be indistinguishable from words ending in o and u, respectively.1.2. Vowels – The Klingon Dictionary – by Marc Okrand
This implies that o is a diphthong ending in a /w/. Since /ow/ is similar enough to /oʊ/, I’ll assume that pronunciation from now on.
|Onset||y /j/ Glide||w /w/ Glide|
|I /ɪ/||Iy /ɪj/||Iw /ɪw/|
|e /ɛ/||ey /ɛj/||ew /ɛw/|
|a /a/||ay /aj/||aw /aw/|
|/o/||oy /oj/||o /ow/|
|u /u/||uy /uj/||Not allowed|
Another issue is that he describes Iy as being pronounced like the ey in key. In American English, this is pronounced /ki/, which suggests that Iy is pronounced like the monophthong /i/ and not the diphthong /ɪj/ like the romanisation suggests. I suppose you could pronounce key /kɪj/, but you would sound a bit weird. This confusion could also be resolved if he provided some IPA transcriptions.
TKD defines the rules for determining stress quite well. A stressed syllable has a higher pitch and a stronger airstream. Stress falls on the final syllable of the stem. In verbs, the stress may be moved to certain suffixes for emphasis. In nouns, stress will move to a suffix if it ends in a glottal stop. There are also exceptions to these rules, but TKD doesn’t indicate them in any way. It just uses the ‘this is not yet understood’ excuse.
TKD does not define phonotactics in any way. Which again is sort of ok considering the target audience of the book, but again it would be nice for it to be in something like an appendix. So I had to deduce these phonotactic rules myself.
The basic syllable structure seems to be CV(w, j, r)(C). A consonant in the onset of a syllable is mandatory. Any consonant can occur in the onset or coda. No consonant clusters are allowed in the onset, but a limited set of clusters appear to be allowed in the coda: w’, y’ and rgh. I think this syllable structure works quite well as it gives Klingon its iconic ‘harsh’ sound.
Word final glottal stops are apparently followed by an ‘echo’ of the preceding vowel, which is short and voiceless. So the name for the Klingon opera ‘u’ (which means ‘universe’ by the way) is pronounced /ˈʔuʔŭ̥/. This interesting quirk helps make the final glottal stop audible and distinct from an open syllable.
I won’t go on about the terrible romanisation system as I covered it in my previous post. But I would like to mention a couple of things that I’ve found out about since reading TKD.
First of all, TKD gives the following description for the pronunciation of I:
I As English i for misfit. Once in a while it is pronounced like i in zucchini, but this is very rare and is not exactly what circumstances account for it.1.2. Vowels – The Klingon Dictionary – by Marc Okrand
This implies that the I functions exactly as the i does in English. So what’s the point of capitalising it? This the worst thing about the romanisation system. Not only the fact that I can be confused with l (which can only be distinguished in serif fonts), but there’s absolutely no reason for it.
The other thing I like to mention is that Marc Okrand clearly states the purpose of this romanisation system:
The system of writing Klingon used in this dictionary has been developed so people who already know how to read English will have a minimum of difficulty approximating the sounds of Klingon words and sentences.1. The Sounds of Klingon – The Klingon Dictionary – by Marc Okrand
Here it clearly states that this system was designed to be intuitive for English speakers. I’m sorry but kind of petaQ thinks this system is intuitive. In fact, there is an appendix in TKD that lists some key phrases, in which a ‘pronunciation guide’ is given for each phrase. This ‘pronunciation guide’, while not perfect, is slightly more intuitive than the official romanisation system. Also, this would also be unnecessary if he got the romanisation system right in the first place.
I would also like to mention that there are some additional capital letters used for other dialects of Klingon. For instance in the Krotmag dialect, voiced stops become nasals in the same place of articulation. This means that b is simply pronounced like a m. However, D is pronounced as a retroflex nasal /ɳ/ and is still phonemically distinct from n /n/, so /ɳ/ is romanisated N. Also in the Morskan dialect H /x/ becomes /h/ in syllables onsets and is dropped in the codas. But because of another sound change in this dialect, Q /q͡χ/ becomes an H /x/, /h/ and /x/ are now phonemically distinct, therefore /h/ is romanised h.
There is lot of interesting vocabulary in Klingon. The KGT discusses semantics in-depth and effectively uses worldbuilding to create believable vocabulary.
For example there are many words for forehead, including tuqvol derived from the word tuq ‘lineage’; and no”och literally ancestor’s tunnel. Both of these words reference how the iconic forehead ridges are used to identify families.
Derivational morphology in Klingon seems to be quite simple. Nouns compounds can be made by adding a modifying noun before the head noun. Also, nouns can be derived from verbs using a suffix -wI’, this basically functions as the -er suffix in English. These processes form a stem, in which additional affixes may be added.
As English is a thing in the Star Trek universe, this means the Klingon has the opportunity to borrow words from English. The word for Earth is tera’ from the word Latin-based word ‘Terra’ and not the more common germanic word ‘Earth’. I guess this is because ‘Terra’ is easier to adapt to Klingon’s phonology than ‘Earth’. I suppose that *‘ergh could have been used, but it isn’t as recognisable.
There are also loan words from other languages in the Star Trek universe. KGT says that qajunpaQ, which means ‘courage’ or ‘audacity’, originally comes a word meaning fiery lava from ‘a language spoken on Krios’. As trivial as this may seem, this adds depth to both the conlang and the worldbuilding, so it feels like I’m reading a description of a real language.
There are plenty of idioms in Klingon that indicate culture such as Hoch nuH qel ‘consider every weapon’, which means to consider every possible solution. But there are also some idioms based on word-play such as quSDaq ba’ ‘sit in a chair’, which means ‘that’s obvious’ and comes from the fact that the words ‘sit’ and ‘obvious’ are homophones (ba’).
Some slang words also have some great world-building integrated into them. For example, ngup ‘authority’ literally means cape, suggesting that authority figures typically wear capes.
Some slang terms are also based on word-play. For instance the words ngIb ‘ankle’ and yeb ‘wrist’ can both be used as insults, because in the Krotmag dialect these words are homophones for yem ‘sin’ and ngIm ‘putrid’
referenceS anD In-jokeS
Some of Klingon’s vocabulary consists of real word references and in-jokes. These words don’t necessary make any sense in context of worldbuilding, but are more like Easter-eggs.
The famous example is the word for fish being ghotI, which is a reference to the joke that ‘fish’ could be spelt ‘ghoti’ since those letters can be pronounced with the phonemes in ‘fish’ (‘gh’ in ‘rough’, ‘o’ in ‘women’ and the ‘ti’ in the suffixes -tion and -tial).
Less of a joke, and more just a reference I have noticed is the word lung ‘dragon’, which resembles the mandarin word for dragon lóng. That might just be a coincidence though.
Klingon appears to be mainly an agglutinative language, which means that some of the grammatical information is expressed through affixes and each affix expresses only one piece of information. Affixes are classified into types that reflect the order they appear in the word. More than one affix of the same type cannot occur in the same word.
Noun morphology is relatively simple. There only suffixes for nouns, which are grouped into five types based on relative order:
- The type-1 suffixes are the augmentative and diminutive. These appear to be an extra bit of derivational morphology since nouns with these suffixes are listed in the dictionary.
- The type-2 suffixes indicate the grammatical number. This is a set of affixes that marks various kinds of plurality (which is optional). For example, there is a suffix, -mey, that implies that the units are scattered. I feel this is quite a unique and interesting approach to grammatical number.
- Type-3 suffixes mark what is referred to as ‘Qualification’, which indicates the speaker’s attitude towards the noun. This is like grammatical mood but marked on a noun rather than the verb, which is interesting. An example of this I found interesting is the suffix -qoq, which indicates that the speaker somehow disagrees with the use of the noun. Basically the equivalent of using ‘air quotes’.
- Type-4 suffixes correspond to English’s possessive determiners (my, your etc.) or the demonstratives (this/that).
- Finally Type-5 suffixes mark case. These are used to mark any syntactic role that isn’t the subject or direct object (which is marked by word order). This group includes a topic marker, –‘e’.
Verb morphology is more complex than nouns. Verbs can take a prefix and up to nine suffixes. For the sake of brevity, I will only cover the ones that I find interesting or important.
Unlike the suffixes, the prefix marks multiple pieces of information. It marks the grammatical person of both the subject and the object. There is also a set of second person suffixes for imperatives.
This prefix behaves more like an affix from a fusional language than an agglutinative one. This may sound bad, but it’s actually a good thing, since this how natural languages tend to behave. For instance, english is usually considered and isolating language, but it has elements of a fusional language.
The chart above lists Klingon’s verb prefixes. You may have noticed that some of these prefixes are the same. This is a feature that is common in natural fusional languages called syncretism.
Another thing I like about these prefixes is how it suggests a change in word order. Klingon is an OVS language, so you would expect the pronoun affixes to reflect this word order. Both subject and object are both expressed with a prefix which suggests that Klingon used to be verb-final.
The asterisks in the chart above indicate reflexives. These are marked with the appropriate intransitive prefix and a Type-1 suffix. Type-1 suffixes include a reflexive, –‘egh, and a reciprocal (for plurals only), –chuq.
There is only one Type-8 suffix, –neS. This is an honorific marker normally used when talking to a superior. I like this because it suggests a society that values hierarchy. A particularly interesting use of this suffix is to derive words that refer to some kind of suicide attack: HIvneS is used when the attack is offensive; while HubneS is used when it’s defensive.
Most of the other suffix types express various kinds of aspect and mood. The Type-9 suffixes, like the Type-5 noun suffixes, express syntactic role.
There are also an additional set of suffixes that are coined ‘rovers’. These are basically suffixes for suffixes, as they modify the suffix that precedes it. Most of these are negatives, with the exception of –qu’ which emphasises the preceding suffix. Weirdly, two of these suffixes do not have this movable property. –Qo’, a negative for imperatives, which occurs between Type-8 and Type-9, while -Ha’, ‘undo’, is used straight after the stem before the Type-1 suffix. While I understand wanting to group the negatives together, I think that classifying these suffixes as ‘rovers’ is very confusing. He should have put these suffixes in their own categories, rather than blaming this bizarre categorisation choice on Klingon linguists (yes, that’s the excuse he gives for this).
Another interesting property of verbs is that they can carry the function of adjectives. In other words, there are no adjectives instead there are verbs listed in the dictionary as ‘to be X’, so they’re like an adjective with a built-in copula.
Klingon is often mocked for having a simple base-10 numbering system, which is not good for a ‘alien’ language. However, this is a bit of an oversimplification. TKD states that Klingon used to have a base-3 system, but was replaced with base-10 in order to understand scientific data from other civilisations (presumably the Federation). The numbers one to ten are:
The numbers one to three are obviously native words from the old base-3 system. I initially assumed that the higher numbers are borrowings, but the KGT actually gives a far more interesting explanation: Numbers four to eight actually come from names for musical tones. This is a very interesting way of deriving names for numbers. KGT states that nine Hut and the suffix for ten -maH are unknown in origin. I would say the simplest explanation for the number nine is that Hut is also from the old base-3 system, as nine is 32 it may of had its own root.
After reading TKD, I was quite disappointed by this number system. However, the explanation into the history of this number system in KGT definitely made up for it.
The basic word order of Klingon is object-verb-subject (OVS). This was intentionally chosen to make Klingon seem alien, as this word order is extremely rare among natural languages.
Subordinate clauses can be formed by using various Type-9 verb suffixes. For example, a clause can be turned into a relative clause by adding the suffix –bogh to the verb.
Yes-no questions can also be formed by placing a Type-9 suffix on the verb. Other questions can be formed with the appropriate question word.
Quite a lot of effort has been put into various registers and dialects, especially in KGT. I have already mentioned some features of somes of the dialects, but I would like quickly mention a particular interesting register called Clipped Klingon .
Clipped Klingon is an abbreviated form of Klingon where grammatical information, such as the verb prefixes, tends to be dropped. It tends to be used in military contexts, as it is quicker to give commands. It is also used in songs to make the lyrics fit in better with the music.
There is a lot more to registers and dialects in Klingon, so I definitely recommend you read ‘Klingon for the Galactic Traveler’ if you want to know more.
TKD mentions that Klingon’s native writing system is called pIqaD, but doesn’t give any in-depth description of it. I don’t think that Marc Okrand is responsible for designing this writing system, as it may have existed before the creation of spoken Klingon for decorating sets.
That being said, this is a terrible writing system. It’s an alphabet with a letter for each phoneme. There are no digraphs and no spelling rules, which not good for a naturalistic conlang. It is also written left to right, just like Latin script.
However, I do like the aesthetics of it. The harsh angular letters fit in with the Klingon culture and the sound of the language. So I suppose this works as a surface level aesthetic element for the backdrops in a Star Trek episode.
To summarise, I will give a pros and cons list followed by a star rating:
- Good presentation in both TKD and KGT.
- Good detailed work on dialects and registers in KGT.
- Good use of worldbuilding.
- Interesting vocabulary with in-depth semantics.
- Phonology sounds unique and harsh as intended.
- Interesting use of agglutinative morphology.
- No IPA.
- Probably the worst romanisation system I’ve ever seen.
- Lack of allophones.
- The vowel system is bland.
- While there are some irregularities, most affixes have only one form. There is only one verb conjugation for the prefixes.
- The writing system is dull.
Overall I think that Klingon is not as bad as I thought it would be. It has some interesting features and I like the way it’s presented in TKD and KGT. I highly recommend reading those books if you’re into Star Trek, as they are a joy to read. However, this is a conlang review and not a book review, and the conlang isn’t particularly great. It’s not completely incompetent, but I think it falls on the bad side of conlanging. Unfortunately, I have to give this conlang 4/10 stars.
Do agree with my opinion? If not, leave an angry comment below and call me a petaQ. Also, I plan on doing more conlang reviews in future so feel free to leave me any suggestions below.
[…] Marc Okrand’s ‘The Klingon Dictionary’ and ‘Klingon for the Galactic Traveller’ (see my Klingon review). It also implies the Brian Bourque has built up Lortho’s phonotactics organically, adding […]
[…] simple base-10 system at least had the caveat of previously being a base-3 system in-world (see my klingon review for more on that). While I understand that someone may use base-10 in their conlang to keep things […]