Préfixes intensifiants dans les Etymologies

Considérations générales

Les Etymologies (telles que nous les trouvons dans LR et VT45,46) incluent différents exemples de formes qui sont désignées comme "intensifiées" par Tolkien, et parmi elles, celles qui sont formées par un préfixe. L'entrée A- dans VT45:5 est peut-être la meilleur point de départ pour la classification de celles-ci.

Nous apprenons d'un préfixe intensifié, distinct par l'origine, bien que similaire en fonction à la voyelle basique préfixée (...) Le préfixe a- était habituellement combiné à l'allongement dynamique de la consonne initiale originale (...) et il peut être appliqué à des mots complètement formés.

Nous pouvons déduire des bribes d'information à partir de ce texte:

Dans ce qui suit, nous explorons ce qui est connu concernant ces deux préfixes distincts plus en détail.

La voyelle radicale préfixée

Nous trouvons une preuve directe de ce mécanisme dans LR:361, entrée I2 - la traducrion est préfixe intensif où i est la voyelle de base et à partir des exemples ITHIL 'Lune' (THIL, SIL) et INDIS- = ndis- 'mariée' nous apprenons qu'en fait ce préfixe ne semble pas être appliqué aux mots entièrement formés mais aux racines THIL (SIL) ou NDIS. VT45:17 ajoute une information où le préfixe est utilisé avec ou sans consonne suivante fortifiée.

Ce qui est sous-entendu par cette remarque est assez apparent à partir de NDIS- (LR:375) où nous apprenons qu'il y a une renforcement de NIS 'femme'. Ainsi, alors qu'il est possible d'utiliser à la fois la racine préfixée et l'allongement de la première consonne pour une forme intensifiée, ce n'est pas nécessaire.

Ceci est quelque peu contredit par l'entrée E- (VT45:11), également traduit préfixe intendif (...) quand la voyelle basique est e mais avec la remarque additionelle suivie par une consonne fortiée. Nous trouvons l'exemple der, ndere - Enderô ('[viril?] jeune marié') > Ender, surnom de Tulkas. Quoique écrit sans majuscule, cet exemple semble se référer aux élaborations des racines DER 'mâle adulte' (LR:354) et NDER- 'forme renforcée de der homme' (LR:375) en *ENDER parrallèlement à NIS- > NDIS- INDIS, et ndere 'marié' semble être une forme en Elfique Primitif plutôt que du Quenya ou du Noldorin - cela ne convient pas à la phonologie des deux langages.

Aucun préfixe n'est listé pour les voyelles basiques o ou u (ou explicitement pour a en fait), mais nous pouvons déduire leur existence à partir dees racines listées dans les Etymologies (voir aussi l'article d'Helge Fauskangers Elfique Primitis).

LR:348 has ANÁR 'sun' derivative of NAR2 and ANAK cf. NAK 'bite' (...) 'jaw', thus confirming the possible elaboration of a root with stem vowel a, the intensification NAR > ANÁR seems to follow a similar idea as THIL > ITHIL.

ÉNED 'centre' (LR356) connects well with NED (...) 'centre, middle' (LR:376), although no intensification as such is apparent from the translation. Note that we don't find a fortification of the first consonant in this example.

ÓLOS 'dream' (LR:379) is an elaboration of LOS 'sleep' (LR:370) - though no direct intensification is apparent from the translation, the idea to get from the general to the more particular connected with the general reminds of 'man > bridegroom' or 'woman > bride'.

No example with a stem vowel u can be found in the Etymologies, but in WJ:389 we see uruk- as a variant of RUKU, indicating that probably u had no special role as such.

All in all, the idea that this particular formation pattern is mostly relevant for roots and not so much for fully formed words seems to work out, though. There is no clear line as to what the meaning of the intensification is supposed to be - the most common outcome seems to be 'a particular out of the general class of objects denoted by the non-intensified form', but even that is not always true.

The A- and N- intensifying prefixes

Most evidence for the second intensification mechanism is listed in VT45:5 entry A- and VT46:6 entry N-. The two entries apparently refer to the same final prefix, and in fact the entry N- includes a reference to the a- prefix, but no reference to N- is given under A-. While this would suggest that the A- entry reflects Tolkien's earlier ideas, the tentative chronology of the entries outlined by Christopher Tolkien in LR:344 would suggest that A- reflects in fact Tolkien's later ideas on the subject (and we will provide supporting evidence for this interpretation from post-Etymologies sources later).

Under N-, Tolkien describes for Quenya an intensive prefix as a blend of a- prefix + dynamic lengthening and a base vowel + nasal initials and of the syllabic forms of nasal initials as mbar. The resulting prefix takes the shape an before vowels (rare); um before p, q, v (umb); an before t, l (=d, and), r (arr), l (all), s (ass), w (anw); before k,g and hence = before vowel (ink, ing).

The entry l (=d) presumably indicates that when a Quenya word-initial l is derived from a root with D- that this consonant reappears when the prefix is added.

In addition, the entry provides the examples umpano 'build', inkale 'radiance', 'Sun', antara 'very lofty', Antaro a name and unquale 'agony'. Consiering e.g. ampano and the root PAN 'place, set, fix in place' (LR:380) we see that this is indeed distinct from a prefixed base vowel which would yield *APAN (see above). Thus, the vowel of the intensive prefix depends on the initial consonant of the base word.

In Noldorin the prefix appears as a with dynamic lengthening afarch, angol (aññol), referring to the unintensified park, ñol. The first can be found under the entry PÁRAK (LR:380) meaning 'dry', the second under ÑOL 'smell', angol is glossed 'stench' (LR:378). We will come back to the meaning of the dynamic lengthening shortly. Note that angol again shows development different from a prefixed base vowel - the root vowel is o and yet an a is used as prefix.

We find a (possible) first emendation in VT46:6 where we learn that aññol, angol appear to have had their initial vowel altered, possibly to ongol and *oññol respectively. Thus, for a moment Tolkien seems to have blurred the distinction even more and introduced a system in Noldorin in which the base vowel was prefixed, combined with a dynamical lengthening of the root vowel.

However, if we look at entry A- (which according to Christopher Tolkien's chronology would be the latest), aññol reappears, so this probably was a transient idea.

Here we learn that the prefix is indeed a- (in Noldorin, Quenya is treated separately) and we learn how the dynamical lengthening proceeds, i.e. parkâ 'dry' ON parkha N parch but apparkâ 'arid, very dry'; ON appharkha, N afarch.

Thus, dynamic lengthening seems to imply that the consonant is doubled in the CE evolution stage and the subsequent phonetical developments are carried through to Noldorin. We may thus e.g. expect c- > ach-, t- > ath-, p > aph- > af-, s- > ass-, n- > ann-, m > amm- and possibly g- > ang-, b- > amb-, d- > and-. Owing to the relation with N-, the prefix would conceiveably become an- before vowels. Note that there is likely a similarity to the effects of nasal mutation for a prefix *an-.

The Quenya evolution likewise undergoes a revision - the prefix is now said to be an, am, añ a generalized form. We see this in action in amparka 'very dry' where the historical form *apparka is rejected, antara 'very high' and ancale 'radiance'. The last example is especially interesting, because it confirms that Tolkien abandoned the concept of having a different vowel dependent on the initial consonant, it apparently replaces incale.

There are still some traces left, though - we learn that un, um appear before q, p, b, but owing to the influence of û- (...) used only in an evil sense , as unquale 'agony., a strengthened form to the root KWAL- 'die in pain' (LR:366).

Intensification in post-Etymologies sources

As far as we can see, Tolkien seems to have stuck with prefixing the base vowel to a CE root vs. an intensive prefix a(n,m,ñ). The clearest example is Quenya ancalima (LOTR, Letters 278-279) 'exceedingly bright' - which doesn't come out as **incalima, **akkalima, follows ankale and not inkale and supports the idea that A- represents Tolkien's later decision.

In WJ:415 Tolkien makes an interesting distinction between an inflectional prefix seen in ekwê which is said to be a primitive past tense marked by the 'augment' or reduplicated base vowel and the long stem vowel, cf. akâra 'made, did' (ibid.) and an extension or intensification seen in akwâ (from kwâ). Like the intensifying prefix a-, a past tense prefix appears to be an object that is only applied to fully formed words, not CE roots.

Sindarin and Quenya estel 'hope' are said to be derived from a prefixed stem vowel (sundóma) from stel yielding e.g. S: thel 'intend, mean, purpose, resolve, will' - where we again observe the action of the stem vowel prefix.

In PM:358 we find Aran Einior, translated 'the Elder King'. This can conceiveably be derived from OS: anjâra with subsequent i-affection of the prefix (which at times comes out at a > ei, cf. the plural pattern alph > eilph (UT:265) and would hence involve a prefix an.

The same prefix may be seen in Cuio i Pheriain anann! (LOTR) 'may the Halflings live long!' (Letters:308) - if one doesn't interpret anann as dative an + and 'for a long time' but as intensified form anann 'exceedingly long'. Neither interpretation is backed up by the actual translation.

WJ:311,315 has the forms thent/estent, possibly meaning 'short/?very short'. If the latter is indeed an intensified form, it clearly is interesting. Should we assume that it simply reflects the base vowel e as a prefix, leading to a re-emerging of the initial root consonants st-? That, however, is exeedingly unlikely. Since the root is STINTA, the base vowel would be i and the historical correct form would then be *istinta > istent (a-affection doesn't occur throughout the whole word, cf. silimâ > silef not **selef (LR:385)). And if one assumes that the form represents analogical leveling, it would be very odd to see the root consonants re-emerging but not the root vowel.

On the other hand, an a- prefix might be i-affected by the base vowel before the base vowel is changed by a-affection, hence *astinta > *estinta > estent. But that would assume that i-affection and a-affection did not act in the usual order. Thus, it is possible that we see here again the intensifying prefix from the Etymologies, but all in all, no explanation seems entirely satisfactory.

Acknowledgements

I would like to thank Carl F. Hostetter and Florian 'Lothenon' Dombach for interesting discussions and valuable comments leading to this article.


Thorsten Renk
thorsten@sindarin.de

Back to the index