Phonology, phonetics and orthography of Nûrlâm

Phonology

Vowels

Black Speech has very few vowel phonemes:

Front Central Back
Close i, iː u, uː
Mid o, oː
Open a, aː

Each vowel can be short or long despite the stress (specially in compound words like Nazgûl). Length of the vowel changes the meaning of word and were used for word derivation in other dialects, but Nûrlâm tries to avoid such situations, like “burzum” – “bûrz” pair with very similar meaning. Long vowels have special characters in scripts of Middle-Earth. In Latin transliteration they are indicated with circumflex sign (^), acute / accent mark (ˊ) or by doubling the letter. For example û, ú and uu indicate the same sound [uː]. In LOTR TT Book Four, Chapter X: “The Choices of Master Samwise”, both û and ú occur in one sentence in words Nazgûl and Lugbúrz. According to Appendix E of LOTR, letters with circumflex should be pronounced longer than letters with acute, however this remark is about Sindarin, and long vowels of non-elven languages usually transliterated with circumflex mark. In later (post-LOTR) notes “Lugbúrz” spelling was corrected to “Lugbûrz”, so it's safe to assume they represent the same sound.

Vowel o is considered rare in Classical Black Speech and often changed to u in borrowed words (i.e. urun from Quenya oron).

Mid front vowel e appears only in borrowed words (usually names) and in other dialects of Black Speech (namely Svartiska). It's usually replaced by i in borrowed words. Some people argue that e should exist in Black Speech because Tolkien never stated that it is absent or rare like o and that it occurs in orc's name Balcmeg. But this name was not included in Silmarillion, is from the First Age when Black Speech didn't exist yet, and it was clearly stated that it is a translation into Gnomish.

Subdialects

Some people assume that letter û indicates rounded close front sound [y] (because of transformation Saruman's moniker “Sharkû” into “Sharkey”). Moreover these two sounds are sometimes mixed in some Tengwar inscriptions. But such proposition is doubtful as circumflex indicates just the length of vowels in other Tolkien's languages too. However if you want to create a subdialect of Nûrlâm with such feature, other long vowels should be transformed too, e.g. î often used to substitute e into [ə] or [ɜː], â into [æ], ô into [øː]. Subdialect may even include pair e ([ə]) and ê ([ɜː]). Some sub-dialects may differentiate between vowels with acute and circumflex according to Sindarin or their own rules (as in MERP), but Nûrlâm's dictionary has all long vowels normalized to circumflex diacritics. The vowel i may be realized as near-close [ɪ] or even as close central [ɨ] in some circumstances to avoid palatalization.

Diphtongs

Diphtongs ai [aj] (“skai”), au [aw] (“Mauhúr”) and oi [oj] (“Dushgoi”) are attested in Tolkien's sources.

Consonants

Labial Dental Palatal Velar / Uvular Glottal
Nasal m n [ŋ] (ng, nk, nd, nt)1)
Plosive (stops) voiceless p t k [ʔ] 2)
voiced b d g
Fricative sibilant voiceless s [ʃ] (sh)
voiced z [ʒ] (zh)
non-sibilant voiceless f [θ] (th) [x] (kh) h
voiced [β] (bh) [ð] (dh) [ɣ] (gh)
Approximant (glides) labial l 3) [w] 4)
plain [j] 5)
Rhotic r 6)
  • consonants are never palatalized (“softened”) before i;
  • ng is pronounced as two separate sounds (as they usually belong to different syllables, or g is followed by h which produces different sound) except at the end of word's root (i.e. “rong” = to dig);
  • p can be aspirated [pʰ] as in English;
  • bh can be pronounced as aspirated or “breathy-voiced” b ([bʰ] or [bʱ] respectively), or as labial approximant/fricative [β]7) (something between [b], [v] and [w]), but only at start or end of the word (like in “bhog” = “good”). More often bh is pronounced as two distinct sounds, specially in words like “bûbhosh” where b and h belong to different syllables;
  • glottal stop [ʔ] appears in colloquial speech separating affixes and some compound words. It can also be produced by h after b (see comment above);
  • voiced labial fricative v is absent in Nûrlam, but exist in other dialects of Black Speech. In borrowed words it is usually replaced by f or sometimes bh;
  • [j] appears only in diphthongs ai and oi and few Quenya-borrowed words starting with y+vowel;
  • r may be pronounced differently: in Ring Inscription it is spelled with tengwa “óre” before consonants or word-finally (“weak”, voiced uvular fricative [ʁ] or alveolar approximant [ɹ]), and with “rómen” word-initially before vowels or between other consonant and vowel (voiced uvular thrill [ʀ] or voiced alveolar thrill [r] – “full” or “rolled” R). This distinction in pronunciation disappeared even in spoken Quenya, so different letters for Black Speech may be just a sign of Sauron's education. Appendix E: “The Orcs, and some Dwarves, are said to have used a back or uvular r”.

Shifting between voiced and voiceless consonants is often used for word derivation (ghâsh = fire, gash = heat, khash = warmth)

Nûrlâm suggests that letters th, dh, kh, gh may be pronounced as aspirated [tʰ], [dʰ], [kʰ], [gʰ] in some dialects, as in David Salo's work and early Adragoor's Svartiska (prior to 2000, [tʰ] and [dʰ] were later changed into [θ] and [ð]). Moreover in Black Speech Tengwar mode these digraphs are written as aspirated fricatives after all short vowels and long ô and û (so thrak is pronounced as [θrak], but Othrod is written as [otʰrod] in Tengwar). However z and sh (and probably s and zh as well) also have distinct glyphs after vowels without changing their phonetical value at all.

Phonotactics

This part of article is not finished and may be changed in the future

revisit possible sound combinations after the dictionary will be expanded

Phonotactics is the part of phonology that studies restrictions on using phonemes. Verb stems are allowed to end with consonant only. Other parts of speech may end also with vowels or diphtongs.

The following consonant clusters are allowed at the start of the stem:

  • Tolkien sources: gl-, gr-, kr-, sk-, sn-, thr-;
  • Fan additions8): bhr-, bl-, br-, dgh-, dhl-, dhr-, dl-, dr-, dv-, fh-, fl-, fr-, ghl-, ghr-, gn-, hl-, hn-, hr-, hs-, ht-, khl-, khr-, kl-, mb-, ng-, pl-, plsh-, pr-, shm-, shn-, shr-, skr-, sl-, sm-, sr-, st-, str-, sz-, thl-, tr-, vr-, zg-, zn-, zr-;
  • Nûrlâm additions: mr-, shk-, shl-;
  • Theoretically possible, but not attested: gm-, hm-, hp-, shp-, sht-, tl-, zb-, zd-, zhd-, zhl-, zhm-, zhr-, zl-, zm-;

The following consonant clusters are allowed at the end of the stem:

  • Tolkien sources: -lg, -mb, -mp, -nk, -rb, -rk, -rn, -rz, -zg;
  • Tolkien sources (unsure)9): -db, -gl, -gr, -lf, -ng, -rg, -shd;
  • Fan additions: -bd, -bdh, -bn, -br, -bz, -dg, -dhl, -dhm, -dhn, -dhp, -dm, -dr, -dz, -dzh, -fr, -ft, -gb, -gd, -ghl, -ghn, -ghr, -ghsh, -gn, -gsh, -gtr, -gz, -hd, -hr, -ht, -kg, -khb, -khl, -kht, -khth, -kl, -km, -kr, -ks, -kth, -lb, -lbh, -ld, -lgh, -lk, -lksh, -lm, -lmg, -ln, -lp, -ls, -lt, -ltz, -lv, -lz, -mbr, -mg, -ms, -mzh, -nd, -ndg, -ndr, -ndsh, -ngh, -nghl, -nr, -nsh, -nt, -nz, -pn, -ps, -psh, -rbh, -rd, -rdh, -rdrg, -rdt, -rf, -rft, -rgh, -rgz, -rkh, -rksh, -rl, -rlg, -rm, -rsh, -rsk, -rt, -rth, -rv, -sg, -shb, -shg, -shk, -shn, -sht, -sk, -sl, -sm, -sn, -sp, -sr, -st, -tg, -thg, -thl, -thr, -tl, -tp, -zb, -zd, -zhb, -zhd, -zl, -zn, -zt;
  • Nûrlâm additions: -kt, -lkh, -md, -mk, -msh, -mz, -nb, -nf, -rzh, -tr, -zhk;
  • Theoretically possible, but not attested: -bg, -bzh, -dl, -dn, -gzh, -fm, -fn, -fp, -fk, -fs, -fsh, -fl, -kf, -khr, -kn, -kp, -ksh, -lf, -lsh, -mf, -mt, -np, -ns, -nzh, -pk, -pr, -pt, -sf, -shl, -shm, -shp, -shr, -zhg, -zhm, -zhn, -zhr, -zm, -zr;

Some consonants in final clusters probably were intended as derivational suffixes of unknown meaning like -b in durb-.

Nûrlâm suggests that consonants in cluster should be either voiced or unvoiced (nasals and glides may join any), so clusters like “-tb” should be avoided (in favor of “-db”), however exceptions like “-gsh” exists. Also, only two consonants are allowed in clusters (digraphs and several consonants occured at joint of syllables do not count), with “skr-” and “-rsk” being exceptions.

Please note that bh, dh, gh, kh, sh, th, zh are not counted as consonant clusters but digraphs which make one distinct sound.

Sound change and Alternation

Shift of consonants is sometimes used for word derivation: “ghâsh” (fire), “gash” (heat), “khash” (warmth). Various types of changes in pronunciation and orthography of word may appear in syntactic words (joining two roots to make new word, e.g. “nazg” + “gûl” = “Nazgûl”) or at morphemes boundaries when similar sounds are going to join. These changes are mostly irregular. However Nûrlâm tries to avoid such derivation, as well as changes in vowel length for this purpose. Also, Nûrlâm rarely reflects merging of morphemes occured during regular inflection in writing, e.g. “off the cliff” will be written as “fipbo”, while being pronounced like /fibbo/. But such phonetical alteration may be reflected in compound words spelling.

Phonetical correspondences

Nûrlâm and other Neo-Black Speech dialects have many words borrowed from various Elvish languages, but there is no consistency in phonetical transformation. One of the reason lies within Nûrlâm's self-restriction to avoid omonyms with other dialects. Therefore some words had to be additionally altered to feat language purpose, sacrificing regularity of phonetical correspondences. Also, some conventions were switched during development, e.g. from Shadowlandian's Quenya “e” → Orcish “î” to Tolkien's “e” → “o”.

add examples for most common phonetical correspondences

Syllable structure

Syllable structure is often described by sound's position. Consonants are marked as C and vowels as V. Syllables are split into two groups depending on their structure: open (ending with vowels) and closed (ending with consonants). Distribution of these groups has significant influence of overall sounding of language. For example open syllables are dominating in Japanese, very frequent in Finnish, a little less than half in Germanic languages and may absent at all in some exotic languages. In Black Speech distribution is roughly 66-70% of closed syllables versus 30-33% of open syllables. Consonant clusters may appear both at start and end of syllable.

Typical word roots in Black Speech are C1VC2 and VC1VC2 (with V being the same vowel and each C being a consonant or cluster of them). Syllables rarely have consonant clusters on both ends. Majority of suffixes follow the pattern VC, and prefixes are CV.

Stress

Most words consist of just one syllable. Otherwise the stress is always placed on the first syllable of the root in Black Speech. This rule also affects compound words (Nazgûl). In Nûrlam main stress is shifted when adding clitics to the end of the word. First syllable remains under secondary stress. In example burzum (the darkness) + ishi (in) = burzumishi (in the darkness). Joining a clitic to the beginning of the word (dagimb = I find), adding prefix (like narish = ally), derivational (for example burzum = darkness) or inflectional suffix (i.e. globûrz = filthy) doesn't change the stress position. These rules allow to resolve possible ambiguity of some clitic with grammar function and compound word.

Intonation

Basing on citation from Appendix F to LOTR about language of orcs “dreary and repetitive with hatred and contempt, too long removed from good to retain even verbal vigour, save in the ears of those to whom only the squalid sounds strong”, it's assumed in Nûrlâm that orcs do not significantly alter tone and pitch inside one sentence for emphasis. So orcs usually speak in “shouting” tone, even in questions (especially on question particle “mar”). Sauron and his high-level servants were probably capable of more variative speech, but they use other languages when it comes to mind control, seduction and corruption.


Orthography

As the preferred method of writing Black Speech isn't known, usually romanization is used. Ring Inscription was written in special mode of Tengwar (with signs for o and u swapped). According to Appendix E of LOTR “The Cirth in their older and simpler form spread eastward in the Second Age, and became known to many peoples, to Men and Dwarves, and even to Orcs”. Also some fictional scripts (usually based on aforementioned two) are used in other dialects (e.g. Maushur in Zhâburi).

Hyphenation

Hyphen appears only in Latin transcription. Usually it separates suffixes from the root and from each other in texts for studying. However in scripts of Middle-Earth suffixes are written merged in one word. Hyphen can be used in real texts to separate clitics or compound words (like bûb-hosh) and avoid confusion with digraphs (like in word “Uruk-hai”, “k” and “h” are separate sounds, not merged into “kh”) and indicate glottal stop which appears in such situations. It may be similarily used with vowels (like dro-ukh = “go forward”).

Concatenation vs. Separate spelling

Usually various grammatical clitics (e.g. verb's person) join the main word through concatenation in spelling, but by Nûrlâm convention they are written separately as stand-alone words when phonemes/words boundaries otherwise may produce:

  • two consecutive vowels;
  • long, hard to pronounce consonant clusters;
  • digraphs.

However, when written in Tengwar, all words will be written without spaces anyway. But Latin separate spelling will ensure that no sound combination into a different character will occur in Tengwar transcribers.

Please note: these rules apply only to word inflection. During the derivation of new words adjacent sounds are often merged into one phoneme or transformed.


1)
only at the end of root
2)
glottal stop
3)
“English dark L”, velarized dental/alveolar lateral approximant [ɫ]
4)
only in diphthong au and Quenya-loaned words containing qu
5)
letter y before vowels and i at the end of diphthongs
6)
“French guttural R”, usually voiced uvular fricative [ʁ] or approximant
7)
according to Édouard Kloczko, “Dictionnaire des langues des Hobbits, des Nains, des Orques et autres créatures de la Terre du Milieu, de Númenor et d'Aman” (Encyclopédie de la Terre du Milieu, volume 4), Arda, 2002.
8)
striked out entries are present in other dialects, but deprecated in Nûrlâm
9)
depends on the way of splitting poly-syllable words without clear etymology
phonology.txt · Last modified: 2022/11/07 00:59 (external edit)