Tia iqet mestani calichniye! Mia James yalun ye, e ninte saana misain ye. Welcome everyone to this lesson! I’m James, and I will be your teacher. In this thread I will be taking a critical look at the Old Tongue and how we can use it as a functional language instead of a smattering of vocabulary.
Some of this will rely on conjecture. These points of conjecture are not intended to claim status as canon from the Wheel of Time franchise. Old Tongue grammar is not a fixed and known entity, though there are many solid concepts. There are some holes in the grammatical and syntactical structure of Old Tongue as it exists now, and this is simply an attempt to fill those holes. It is not the only method of doing so, simply the method that this poster personally feels best fit the circumstances surrounding the information given, as well as the various sources from which Robert Jordan drew inspiration. As much as possible, the original works will be used as the sole sources. No official party associated with the Wheel of Time franchise, be it Robert Jordan, the Jordan Estate, Brandon Sanderson or Tor Publishing, endorses this method or the viewpoints expressed herein.
This thread will take a look at what we know to be elements of the Old Tongue and attempt to reorganize them from the ground up. A great portion of information will be synthesized from the books directly or from the Wheel of Time Companion, with additional assistance taken from this post on the Wheel of Time wiki.
< • >
The Problem with Pronunciation
One of the standing issues right off the bat is how to pronounce a word on sight unseen. Looking at the given pronunciation in the appendices of the books, the overall model of Old Tongue phonemes (or “sounds in the language”) appear to line up with those phonemes present in English.1 There are very few nasals (apart from the sounds /m/ and /n/ as appear in “many”), no apparent trills or taps (/ʙ/, /r/, /ɾ/; /ʙ/ is the sound of buzzing lips, /r/ as in the Spanish word “perro,” and /ɾ/ as in the Spanish word “pero”), and an abundance of fricatives (/f/, /v/, /s/, /z/ as appear in “flavor,” “say,” and “zoo”) and plosives (/p/, /b/, /t/, /d/, /k/, /g/ as appear in “pub,” “toad,” and “keg”). There are also similar consonant clusters present in words containing “ch” and “j” (/t͡ʃ/ and /dʒ/ as appear in “cheese” and “judge”). Individual vowels in Old Tongue also line up with English vowels, more or less, with a more prominent preference for front-sounded vowels.
Words containing “c,” “q,” and any potential diphthongs (vowel clusters) are the largest source of ambiguity. In most cases, the use of “c” serves as an orthographical misdirection – the use of “c” in English could denote either a /k/ or /s/ sound. However, usage in Old Tongue seems to be unilaterally consistent; if by itself “c” will indicate a /k/ sound (ascar, cadin’sor, caldazar, car’a’carn, olcam, scrup), while if paired with -h it will indicate a /t͡ʃ/ sound (airach, rare cluster). There are still moments of confusion, however, as with the word orcel.2 Interestingly, the cluster “ch” appears to always be a /t͡ʃ/ sound instead of the more guttural /x/ (as in “loch” or “ach”). The letter “q” is more complicated; if paired with -u it seems to follow the English pronunciation of /kw/, but frequently it appears followed by another vowel entirely. It appears to take on a /k/ sound when appearing by itself in such a manner (raqit, qaiset, qen).
Combinations of vowels in Old Tongue words may or may not be represented by diphthongs (two vowels sounding together), but almost universally all representations of doubled vowels condense into monophthongs. They have the following phonological sounds attached to them:
• “aa” → /a/
• “ai” → /ai/, /e/
• “ae” → /e/
• “ei” → /ei/, /e/, /ai/
• “ie” → /i/
• “oo” → /u/
• “ou” → /u/
At no time would a doubled vowel (“aa”, “oo”) be voiced as separate sounds, unless the word already contains a full stop (for instance, Tuatha’an /tu’aθəʔan/).
In short: sounds in Old Tongue model sounds in English; “c” sounds as a /k/ when by itself and /t͡ʃ/ when paired as “ch”, and “q” sounds as a /k/ when not succeeded by a “u”; and vowels are similar to those found in English with the above exceptions.
Making Sense of Morphology
When looking at Old Tongue, there are a variety of methods by which one can pluralize a noun. Some methods directly contradict each other, while others appear to be standalone treatments for specific words. The methods are:
• Add “-i” (shar “blood” → shari “bloods”)
• Add “-a” (sei “eye” → seia “eyes”)
• Remove a letter (dareis “spear” → darei “spears”)
• Add “-n” (tai “true” → tain “true”)
• Add glottal stop, double vowel if and only if word ends in “-n” (athan “person” → atha’an “people”)
The fact that there are so many methods of pluralizing a noun suggests that Old Tongue is a context-sensitive language, as seen by the inherent contradictions in some points. However, it could also mean that there are certain exceptions to whatever rule governs this pluralization, indicating unseen etymological history within Old Tongue. One can see a pattern emerge in the larger scope of nouns and pronouns; therefore, one may assume a general rule as follows:
[noun] + i (+ a)
where an -i suffix may be attached if a noun ends in a consonant and an -a suffix attached if a noun ends in -i.
This process of suffix modification seems to carry over to verbs, as is evident in the inscription on the Horn of Valere. Looking at the phrase Tia mi aven Moridin isainde vadin, the verb “to be” is contracted with the adverb “not,” allowing for the compound isain + inde → isainde. In addition, the verbs ursta, nolve, and gemarise (fix, give, make) morph into urstae, nolvae, and gemarisae to indicate the past tense. This suggests that verb modification also takes place as a suffix, and that such is the grammatically correct placement for word modifiers.
In Old Tongue, there is also the concept of compound nouns, or phrases put together to make what could be termed proper nouns. Thus, we have phrases like Tain Shari or “true bloods,” as well as Sha’mad Conde or “thunder walkers.” This given structure is inverted in the similarly built Shae’en M’taal or [lit.] “dogs of stone,” and both follow what would be considered proper English grammar in their construction. Inclusion of a prepositional phrase seems to invert the structure of these compound words. It can therefore be stated that when creating a compound word the following rule is followed:
[adjective/modifier] + [noun]
[noun] + [preposition] + [adjective/modifier]
If, however, the word is a contraction of two other words, the reverse holds true. Siswai’aman literally translates to “spears-dragon,” even though it is interpreted as “spears of the dragon.” Another example would be tai’daishar, meaning “lord of glory” even though it translates to “lord-glory.” In a similar sense there exists the compound caba’donde, or “a horse to ride.” This is a literal translation; caba means “horse” and donde is the verb “to ride.” For prepositions, there is also an exception when dealing with far or “of.” This means we have Far Dareis Mai “of the spear maidens” and Far Aldazar Din “of the eagle brothers.” Taking all this into account, the above rules can be amended with their relative exceptions thusly:
[adjective/modifier] + [noun]
[noun] + ’ + [verb/adjective/modifier]
[noun] + [preposition] + [adjective/modifier]
Far + [adjective/modifier] + [noun]
One benefit of Old Tongue over several languages is the general lack of articles attached to the nouns (the house, a tree, etc.). In general practice articles are implied without statement, with the only exceptions being pronouns (souvraya “my mind”) or an explicit quantity of things (choryat caba “five horses”).
The two biggest roadblocks to utilizing Old Tongue as a method for parlance center around the verbs, for they are few in number and lacking conjugation. One may assume that, since there are two known related forms of the verb “to be” in misain “[ I ] am” and isain/ain “[he/she/it] is,” it could perhaps rely on similar conjugation to, say, Romance languages, or perhaps even Irish Gaelic. If that is the case, there would be anywhere between 4-6 conjugations per verb. However, the biggest argument against this comes from a comparison of the known vocabulary of Old Tongue to English. In English, to imply different tenses of a verb, modifiers such as adverbs and auxiliary verbs are attached; this changes present-tense “I play” to its negative “I didn’t play,” its preterite/past tense “I played,” its future tense “I will play,” its conditional “I could play,” etc. Including the negating adverb inde, there are several other verb modifiers and mutations seen in the franchise that give rise to different tenses, making it more aligned to English than any other potential language base.
From the utterances Mat provides after his cleansing, we can see that the present tense is simply the verb by itself, in a similar way to English or Cornish.3 The sentence Sene sovya caba’donde ain dovienya, translating to “Luck is a horse to ride like any other,” uses a conjugation of the irregular verb “to be”4 and leaves it at that to form the present tense.
Again, looking at the contraction of isain and inde into isainde we can intuit that negation-form of a verb arises from a similar structure. This will work for sentences using the equivalent of “isn’t,” “don’t,” or “doesn’t;” if one is trying to deny a thing has a quality (Mat’s assertion Inde muagdhe Aes Sedai misain ye, “I am no Aes Sedai meat”), one instead puts the negative inde at the beginning of the sentence.
[verb] + inde
inde [Object] [verb]
The only complete form of inquisition in Old Tongue is given by Birgitte Silverbow at Falme, when she asks Mat Nosane iro gavane domorakoshi, Diyned’d’ma’purvene? (lit. and given translation: “Speak we what language, Sounder of the Horn?”). Here, the structure morphs to VSO or Verb-Subject-Object, while the verb itself remains in its base conjugation. In addition, the adverb “what” is inserted in between the subject and the object, allowing for journalists to ask their five Ws.
[Verb] [subject] [adverb] [Object]
A formalized, and perhaps even florid, example of the declarative can be seen in Mat’s address to Tylin in A Crown of Swords. His statement Deyeniye, dyu ninte concion ca’lyet ye includes the verb lyet combined with an emphatic ca’ implying something stronger than just stating “I come.”
ca’ + [verb]
No sentences in the original text show an example of past tense; however, from the verbs gemarise, nolve, and ursta listed in the Companion we know that the past tense is formed by changing them to gemarisae, nolvae, and urstae. The suffix -ae is also listed as denoting a passive voice in the Companion.
[verb] + ae
Currently there is no method of forming a future tense in Old Tongue. However, one may use the conditional tense as a stand-in for any situation necessitating the future tense (see below).
There are two separate methods of forming the imperative. When one is speaking from a position of authority, as a commander to troops or a person willing some aspect of their life to be so, one changes the sentence structure from Object-Verb-Subject to Subject-Object-Verb. This can be seen from a phrase spoken during Mat’s cleansing, Muad’drin tia dar allende caba’drin rhadiem (lit. “Footmen to forward pass cavalry prepare”), as well as his later utterance during events in The Shadow Rising, Mia dovienya nesodhin soende (lit. “My luck through [this/it] carry;” though no official translation exists this can be surmised by referring to the Companion). If, however, one speaks from a lower position, as a commoner to the Aes Sedai sisters, the standard OVS structure continues to apply. This is seen in Mat’s cleansing by one of his first phrases: Mia ayende, Aes Sedai! “Me release, Aes Sedai!”
[subject] [Object] [Verb]
[Object] [Verb] [subject]
The auxiliary verb punia meaning “may” can be added to sentences to imply conditional tense. In this case, while there is no direct parallel to the conditional tense there are instances of combining verbs. In the sentence Dovie’andi se tovya sagain, we see the closest structural parallel. The verb isain is combined with the noun sag to imply “it is time,” and is preceded by the verb tovya “to roll.” The implication is that word order is directly opposite that appearing in English, and thus it may be inferred that one may add punia after a verb to imply the conditional tense (“I may go” or “I may teach”). Given that there is no known future tense of Old Tongue, this may be the closest approximation to a future tense as exists, with speakers declaring their intent to engage in an action.5
Currently there is no method of forming the imperfect tense in Old Tongue, for use in situations as would be expressed in English by “I am running” or “I was swimming” implying actions as they occur and before their completion. However, it may be possible to approximate the imperfect tense.
From this point on is conjecture. If we assume that verbs in Old Tongue operate in a similar capacity to English, one could form the imperfect tense by adding “to be” in conjunction with the verb. This would allow for a similar understanding as is given in English that the verb is in the process of completion, while allowing for other tenses such as past and conditional to increase variety.
[verb] misain (punia)
Here the misain conjugation of “to be” is speculated as it is listed as an emphatic conjugation. It is also assumed that punia be added after misain in the event of the conditional imperfect tense (“I might be running,” “I might be swimming”).
Conjecture ends here.
Old Tongue is in most cases an OVS language – that is to say, its sentence structure follows an object-verb-subject pattern.6 Speakers of Klingon will recognize this pattern, as it is one of the very few examples of a language that follows such a structure. This means that even before we can take a crack at building sentences we have to re-learn how sentences are built. For example, in the beginning of this post I greeted everyone with Tia iqet mestani calichniye. A literal translation of this would be “To this lesson welcome.” Looking at an example from the books, we have Mat’s exclamations from his Healing in which he exclaims: Mia ayende, Aes Sedai! Caballein misain ye! Inde muagdhe Aes Sedai misain ye! Literally: “Me release, Aes Sedai! [ A ] free man am I! No Aes Sedai meat am I!”
The vast majority of sentences do follow this pattern, though there is one very notable exception to this in the present tense. The inscription on the Horn of Valere reads Tia mi aven Moridin isande vadin. Literally translated, this becomes “To my call death isn’t [ a ] bar.” This seems to suggest an SVO structure, or object-subject-verb structure, can sometimes be employed.7 The most likely explanation for this is inflection of a poetic or dramatic flair, or perhaps to emphasize a particular portion of the sentence.8
Aside from the Horn’s text, there are also exceptions as stated above for grammatical purposes. These appear to follow certain situational instances; however, as with any language it is entirely plausible that one may bend convention in order to better fit a particular need or personal preference.
Putting It All Together
With the information gathered above, sentence structure for statements and imperative statements from a position of submission can be ordered by using the following formula:
[inde] [Object] [prep. phrase] [ca’ ] + [Verb] + [verb modifiers] [aux. verb] [subject]
Questions are formed as this:
[ca’ ] + [Verb] + [verb modifiers] [aux. verb] [subject] [question adverb] [inde] [Object] [prep. phrase]
Imperative statements from a position of authority are formed as this:
[subject] [inde] [Object] [prep phrase] [ca’ ] + [verb] + [verb modifiers] [aux. verb]
Proper nouns can be placed either at the beginning or end of a sentence. General practice seems to put it at the end, as seen in sentences such as Mia ayende, Aes Sedai and Nosane iro gavane domorakoshi, Diyned’d’ma’purvene? However, there can be exceptions as exemplified by Deyeniye, dyu ninte concion ca’lyet ye. These exceptions likely occur as a result of personal taste and/or gravity of a situation.
< • >
Languages are messy. Looking at the Old Tongue is no different – there are exceptions, there are inconsistent models, and in the case of structural integrity whole pieces seem to be absent. But the challenge of using Old Tongue as a spoken language is not impossible, and certainly not outside the realm of the average fan. Using no more than a crucial eye to analyze the sayings of our knavish ta’veren and his interactions with various other characters, we can determine how the Old Tongue could be used as a spoken language without asking a single question of the original master. Once again, even years after his passing, Jordan leaves enough clues for the eagle-eyed to piece together the structure hidden in plain sight.
N’am isain fel domorakoshi. Iqet mestani aris nolve punia.
1 When “English” is used herein, the term will refer to Standard American English (SAE), as opposed to Standard Canadian English, Received Pronunciation, or other dialects. This is based on Robert Jordan’s own linguistic background in using SAE.
2 This is why many conlangs refrain from using the letter “c” unless appearing in the cluster “ch” to form the /t͡ʃ/ sound. The reasoning behind this is that many readers tend to read “c” with English values, morphing it into a /s/ before i/e and into a /k/ elsewhere. A famous example of this is Tolkien’s character Celeborn; the author intended a /k/ sound, yet most readers mispronounce it as Seleborn.
3 Given Jordan’s heavy reliance on Arthurian and Celtic mythos, the most likely linguistic roots in Old Tongue are from those languages spoken by the people from whom the myths originated. Chiefly, these languages are Irish Gaelic, Cornish, and Breton. However, as can be seen by continued reading, the vast majority of necessary verb tenses to hold a structured conversation are already present in the extant works.
4 The verb “to be” is one of the very few verbs in Old Tongue that actually shows conjugation, most of which is situational.
5 There are several real-world precedences for this, surprisingly. A number of languages – including Chinese, German, Japanese, and Norwegian – lack a defined future tense, and context alone informs the practitioners when a sentence is meant to imply some action in the future. There is even a dialect of German spoken in the United States that refuses to use a future tense in favor of a conditional, with cultural implications that one can only intend to do something because they don’t know what the future holds.
6 In this, Old Tongue stands almost alone when compared to sentence structure of languages from around the world. A full 87% of languages follow either an SVO (“She him loves;” Japanese, Latin, Hindi) or SOV (“She loves him;” English, Russian, Mandarin) structure. Only 1% of languages follow this OVS structure, and many of them are minority languages. Among them are Urarina, Huarijio, Hixkaryana, and Apalaí, none of which have more than 5,000 speakers at this time.
7 This phenomenon is present in several languages, most frequently as a means to imply emphasis. It is analogous to what can be seen in English as the difference between “I don’t believe that” and “That I don’t believe.” It can also appear in Norwegian (“Jeg tror ikke det,” “I don’t believe that,” can become “Det tror ikke jeg,” “That I don’t believe”) as well as in Russian (“Я закончил задание,” “I finished the task,” can become “Задание закончил я,” “I finished the task”).
8 It is also theorized that the phrase could have been a product of being written early in the Wheel of Time franchise before the language normalized. If so, a more grammatically correct iteration would be “Tia mi aven vadin isainde Moridin.” (T/l: To my call [ a ] bar is not death.) The author of this thread, however, believes in upholding the original work of the author, and feels that sufficient grounds are present to expect a certain amount of artistic license when constructing sentences.