What do I mean when I say Irregular and Special verbs? Irregular verbs, are verbs that do not conjugate like “normal”. Normal, in this case, being the most common conjugation of a type of verb. There is no painless way to determine whether a verb is irregular or special, so one must therefore memorize these verbs. I should point out that it is not so important to remember how each and every one of these verbs is irregular; but rather, remember the verbs that are irregular and refer to the verb conjugations chart for the appropriate conjugation (this chart can be found in the resources division of the language section …. This lesson is merely an attempt at explaining some of the reasons behind such seeming irregularities. This lesson is not all that relevant to those wishing to be able to write Sindarin. All you really need to do is look at the verb conjugation chart to see what the correct form of the verb should be. However, I include this here because I believe that knowing the reasons for seeming irregularities will help you to develop a stronger sense of the language and how it functions.

*Note: This is quite possibly the most complex lesson of the entire series, so do not be afraid if you do not understand what I am saying. It is 1) not really necessary if you are just interested in learning to write Sindarin and 2) it usually takes a few reads for this sort of complex stuff to sink in. However, I did not remove it from the lessons because I felt that the information was too important to remove.

The following verbs are all irregular or special:

Groga- “feel terror”
Loda- “float”
Toba- “roof, cover over”
Soga- “drink”
Elia- “rain”
Anna- “give”
Drava- “hew”
Thora- “fence”
Banga- “trade”
Nod- “tie, bond”
Tog- “lead, bring”
Gwedh- “bind”
Trenar- “recount, tell to end”
Boe- “it is necessary, one is compelled to, one must”

These verbs can then be broken down into three differing categories: Impersonal verbs, U Surviving before a Nasal, and various Irregulars. We will go over each in turn:


So what exactly is an impersonal verb? An impersonal verb is a verb in which its action does not directly affect a person. In other words, it is an “action” that a person cannot do. For example, I cannot make a sentence such as “Gildor it is necessary” or “Gildor rains”. Both are obviously incorrect. Sindarin has only two known impersonal verbs. These are Elia- “rain” and Boe “it is necessary”. We can easily tell that these are impersonal because it is impossible for a person to “rain” or “be necessary”. Let us first discuss the impersonal verb BOE.

“Boe”- is actually a Noldorin verb that has been “updated” to fit with the phonology (sounds) of mature Sindarin (in Noldorin this verb appears as BUI-). We do not have any attested examples or Boe in use, but it may be possible to use this verb in such sentences as:

Boe maethad in yrch “It is necessary to fight the orcs”
Boe ‘nin edhil maethad in yrch “It is necessary for the elves to fight
the orcs”
Boe anim baded “It is necessary for me to go” = “I must go”

This verb does not appear to be inflected; that is, it does not appear to change in any way. Therefore we do not have any tense changes or pronominal endings.

Note: We use the gerund instead of the infinitive here. This will be explained in greater detail in the lessons on sentence structure.

The other impersonal verb that is attested in Sindarin is Elia- “rain”. To keep things simple and understandable, we will discuss only the forms of this verb that differ from a normal conjugation of an A-stem. The only forms of this verb that do not comply directly with the conjugation of an A-stem are the Past Tense and Present Tense. All other forms of the verb are conjugated as usual.

The Past tense of Elia- seems to have two forms instead of just one. We have the regular “normal” conjugation Eliant “rained”, and we also have an irregular form Aul “rained”. Both seems to exist side by side, with one not being more important that the other.

In the present tense, instead of seeing the expected Elia “rains” we have the irregular form Ail “rains”. This does not seem to exist side by side with the regular form, instead wholly replacing it.

Now I am sure you are asking yourself; “ok, now do I use the irregular, or the regular form of the past tense?” My suggestion would be to use the irregular form. The development of an irregular form of a verb usually tends to “take over” the spot of the older form. However, it is quite possible to use the “regular” form and not be entirely incorrect.


This irregular form of conjugation deals with several mixed conjugation verbs that have been mentioned before. To repeat, these verbs are:

Groga “feel terror”
Loda “float”
Soga “drink”
Toba “cover, roof over”
Nod “tie, bind”
Tog “lead, bring”

These verbs should all be conjugated like normal mixed conjugation verbs, except in the past tense. If you can remember back to the conjugation of I-stems, we talked about a feature known as “nasal infixation”. This was the placement of the “nasal” sound -n- inside the verb itself. If you remember, this caused the consonant directly following the nasal infix to revert to its older form. For example:

ped “speak” > pent “spoke”
dag “slay” > danc “slain”
cab “leap” > camp “leaped”

What we have with these irregular verbs is something very similar. The past tenses of these verbs are also formed with a nasal infix, and cause exactly the same changes in the final consonant, as normal I-stems. However, there is one additional feature that these verbs possess. Instead of just the final consonant being altered to its older form, the vowel before the infix also reverts back to its older form; in this case, -u-. All of these irregular verbs derive from primitive “roots” that contain “u”. The nasal infix “shielded” this original sound so it did not disappear. Thus we have the following 3rd person singular past tense conjugations for these verbs:

Groga “feel terror” > Grunc “felt terror”
Loda “float” > lunt “felt terror”
Soga “drink” > Sunc “drank”
Toba “cover, roof over” > Tump “covered, roofed over”
Nod “tie, bind” > Nunt “tied, bound”
Tog “lead, bring” > Tunc “lead, brought”

For all other persons in the Past Tense, either of two things happens. 1.) Stems that originally have no final vowel conjugate as normal I-Stem verbs with all intervocalic changes and i-umlaut(u>y) or 2.) Stems that originally have a final -a conjugate by adding -e- before the pronominal, triggering intervocalic changes, but no i-umlaut.

These would appear as the following before pronominal endings (remember the clusters nc, nt, mp, and nt cannot occur in the middle of words and are therefore changed!):

Grunc > Grunge- + pronominal ending
Lunt > lunne- + pronominal ending
Sunc > Sunge- + pronominal ending
Tump > Tumme- + pronominal ending
Nunt > Nynni- + pronominal ending
Tunc > Tyngi- + pronominal ending


Along with all of the other irregular verbs, we have a few that are extra-irregular :). These verbs are all conjugated irregularly because of their each “individual history”. Therefore we are left with quite a few little quirks and odd arrangements here and there. For clarities sake, we shall only discuss those features of these verbs that differ from the regular conjugation of the A-stems, or I-stems. Let us start out by listing those verbs that fall into this category:

Anna- “give”
Drava- “hew”
Gwedh- “bind”
Thora- “fence”
Trenar- “recount, tell to the end”
Banga- “trade”

A starting note: Several of the following verbs will have some strange derivations with the diphthong AU. AU sometimes becomes O in several tenses. When this happens, this O becomes impervious to I-umlaut.

Let us begin with the verb Anna- “give”. Anna seems to follow the normal conjugation of an A-stem except in the Past Tense and Part Participle. The Past tense seems to be Aun, which in turn becomes One- before pronominal endings. This is a change of the Diphthong AU to O. For example, we have:

ONEN i estel edain “I gave hope to the Edain” (ONE- + N “I”)

The past participle is ironically exactly similar to the Past Tense form with the pronominal ending “-n”; Onen. However the result is achieved by a different process. The past participle would, as usual have a plural form … “-in”. However O-, in this case does not umlaut. This is because O- that is derived from AU- does not umlaut.

AUN “gave” -> ONE- “gave” + EN > ONEN “given” plural ONIN (not I-umlauted to ENIN, because O comes from AU)

So we get ONIN in the plural past participle (remember that the past participle is the only participle with a plural form!).

The verb Drava is conjugated like a normal A-stem except in the past tense. Here we seem to have an irregular form Dramp, as if this verb were an I-stem. This form seems to be poetic and exists right along side the normal form Dram.

The verb Gwedh seems to be regular, except for the past tense. Tolkien indicated that an irregular form of the past tense Gwedhant came into use, while the regular form of the past tense Gwend came to be considered poetic. The passive participle was most likely also changed from Gwennen to Gwedhannen. This may suggest that the verb Redh- “sow” also underwent a similar change.

Thora- “fence” seems to be quite regular except for the Past Tense, Perfective Active Participle, and the Passive Participle. The past tense of Thora- seems to be Thaur, which becomes Thore- before pronominal endings. This is again, like with the verb Anna, a change from -AU- to -O-. The perfective active participle (remember the PAP is formed by adding -iel to the past tense form of the verb) seems to be Thóriel.

Thaur “fenced”: Thore- “fenced” + iel > Thóriel “fencing” (not Thúriel because O comes from AU)

Normally we would expect this to be Thúriel (We use thore- because we are adding an ending, although not pronominal). However, because this -O- of Thore- is derived from AU it does not change its form. The passive participle also seems to be slightly altered. In its plural form we might expect Therin (from Thore + in). Instead because -O- comes from -AU-, and therefore does not change, we end up with Thorin.

Thoren “fencing” (pp) > Thorin (not Therin because O comes from AU)

The verb Trenar- seems to have some irregular forms because of its descent. Trenar- seems to conjugate like normal except for the Past Tense, Perfective Active Participle, and Passive Participle. The past tense appears to have an irregular form Trenor. This change from A to O probably is because the A was primitively long A. This has some interesting effects and is very similar to the change of AU to O. The Perfective Active Participle is given as Trenóriel instead of the expected Trenúriel. This is probably do to A being long A primitively. Therefore O does not change its form. This is likewise in the plural form of the Passive Participle. Instead of Trenerin we instead see Trenorin.

Trenor “told” + iel > Trenóriel “telling” (pap) (not Trenúriel, because O comes from long A)

Trenor “told” + en > Trenoren “told” (pp) > Trenorin (pl) (instead of Trenerin, because O comes from long A)

The verb Banga- is a rather odd verb. In the Etymologies we are given a form Banc but we are not told what it means. This could be a past tense form of the verb Banga, but it may also be a noun meaning “trade”. If Banc is a form of the verb Banga-, then Banga- is probably conjugated like a normal mixed conjugation verb. However, it may not even be a form of a verb at all, and instead might be a noun. To make this even more confusing, both might be right as well! Banc- may be the past tense form of Banga- and it might be a noun Banc! My suggestion would be to treat Banga- as a mixed conjugation verb, and treat Banc as a noun as well.


• There are three categories of Irregular and special verbs: Impersonal, Original U, and Various others.
• Impersonal verbs are verbs where the verbal action does not directly target a person
• Original U surviving before a nasal are several mixed conjugation verbs that have a nasal infix that not only changes the consonant in front of it to its older form, but also the vowel before it into older u.
• AU seems to be the biggest driving factor with the Various Others.
• AU often becomes O. This O cannot then be I-umlauted.
• O derived from primitive long A, also appears to be impervious to I-umlaut