Latin Stems: Part II

Verbs and Verbal Stems

A Recap

In part one, I introduced at the idea of stems and how it works in Latin nouns and adjectives. We saw that Latin nouns and adjectives can be broken into three parts.

root + stem vowel + case-number affix

By “root”, we meet the meaningful part of any word, which does not change and is found at or close to the beginning of a word. A “case-number” affix is the part that provides grammatical information – that is, what case the word is in, and whether it’s singular or plural. The “stem affix” is something, usually a single vowel, sandwiched between the two.

The combination of the root and the stem affix create what we call the stem of a Latin word. It’s to the stem that the final grammatical stuff is added. The construction process looks a bit like this:

[ [ [ root ] + stem vowel ] + case-number affix ]

The brackets here are to mark the different stages of construction, starting with the basic root. For example,

  • [ [ [ domin- ] -o- ] -s ] -> dominus
  • [ [ [ frūct- ] -u- ] -s ] -> frūctus
  • [ [ [ ign- ] -i- ] -s ] -> ignis
  • [ [ [ di- ] -ē- ] -s ] -> diēs

Breaking up the word into three parts has many uses: the most important is that it cuts down the number of grammatical endings that we need to learn.

But that’s enough about nouns – what about verbs? What do verbal stems look like?

Part II: Verbal Stems

Latin verbs are also built through this three-step process. However, there is one crucial difference: while the stem of a Latin noun or adjective never changes, stems for verbs do. Every Latin verb you will encounter is built around the combination of a root and one of three types of stem.

It is then to these stems that the grammatical information at the end is added, which is expressed in what I call the tense-person affix. This isn’t a great term, as the suffix conveys much more information than just the tense (present, future, etc.) and person (first-person singular, etc.) of the verb, but it will do for now.

But don’t worry! A verb’s stem may change, but it does so in a very consistent way. The goal is this article is to show you how this works. My earnest hope is that it will successfully explain things that would otherwise remain mysterious component parts of Latin verbs, and that it will therefore make verbs easier to understand.

The way to understand this system of changing stems is to first imagine all Latin tenses together in one big set. These include not only what we normally think of as ‘tenses’ (e.g. present, future, imperfect, perfect), but other verb forms too, like infinitives and participles. Such a set would look something like this:

These different forms of a verb are expressed by dedicated verb endings, which are added to the verb’s stem. But which stem? We can divide them into three ‘stem groups’.

Each stem group corresponds to one of three stems.

  1. To make a verb in the present, future or imperfect tenses, we combine the right endings with the present stem.
  2. The perfect, pluperfect and future perfect are formed with endings added to the perfect stem.
  3. Future participles, past participles and the supine require the supine stem.

So, what do each of these stems look like? How are they different from each other, and from the bare root of the word?

The Present Stem

The present stem is very common, founded in many tenses and verb forms. It is also the simplest. Every verb has a present stem and, except in irregular verbs, it is created through the combination of a root (the meaningful part of a verb) and a vowel. The basic structure looks like this, where ‘V’ is a vowel:

[ [ root ] –V– ]

For example, the present stem of the verb moneō, monēre ‘warn’ is monē-.

[ [ mon- ] ē]

The present stem of amō, amāre ‘love’ is amā-.

[ [ am- ] -ā- ]

This is why we find amā– in so many forms of the verb ‘love’, such as amās ‘you love’, amābam ‘I was loving’, amāre ‘to love’ and amandus ‘to be loved’.

Four stem vowels are possible: ā, ē, e and ī. We can therefore sort present stems into five groups:

  1. ā
    For example: amā– ‘love’, parā– ‘prepare’, laudā– ‘praise’
  2. ē
    For example: monē– ‘warn’, habē ‘have’, vidē‘see’
  3. e
    For example: dūce- ‘lead’, lege- ‘choose’, dīce- ‘say’
  4. ī
    For example: audī ‘hear’, dormī ‘sleep’, invenī– ‘discover’
  5. Irregular
    For these, the present stem is more difficult to ascertain. For example, the present stem of esse ‘to be’ is probably just es-, without a vowel, but we find it changed by the addition of various endings.

To the best of my knowledge, there is no single rule that determines which vowel each verb’s present stem uses; it seems random, but at least the present stem of a verb does not change.

These groups may look familiar – they match up to the traditional four conjugations! Verbs that form the present stem with ā belong to the first conjugation, those with ē to the second, and so on. There is a good reason for this. The traditional conjugations were determined by a verb’s ending in the present infinitive (-āre, -ēre, -ere or -īre). Yet, the present infinitive itself is really a combination of the affix -re and the present stem. Therefore, here is a handy rule:

Rule 1: To find the present stem of a verb, take off -re from the present active infinitive. This will be the second principle part of the verb given in a dictionary.

For example, if you look up ‘love’ in the dictionary, it will give you

amō, amāre, amāvī, amātum

Taking off the -re from amāre gives you amā-. Now you have the present stem, you can use it!

As shown in the diagram of tenses above, it is to the present stem that the endings of numerous tenses and verb forms are added. The present, imperfect and future tenses, present participles, present infinitives and gerundives require the present stem.

Here is a selection of forms of moneō, monēre ‘warn’, which all use the present stem monē-:

  • monēs – you warn (present tense)
  • monētis – you all warn (present)
  • monet – she warns (present)
  • monēbāmus – we were warning (imperfect)
  • monēbunt – they will warn (future)
  • monē – I will warn (future)
  • monēre – to warn (infinitive)
  • monendus – to be warned (gerundive)
  • monē – to be warned (infinitive)

Within these , we can see monē– combined with various tense-person affixes, which, importantly, will be the same for other verbs too. For example, -s is a second-person singular affix for verbs of all stem groups – compare monēs, amās, audīs, even irregular es. Applying the same endings used above with monē– to the stem amā– gives us the forms of the verb for ‘love’.

  • amās – you love (present tense)
  • amātis – you all love (present)
  • amat – she loves (present)
  • amābāmus – we were loving (imperfect)
  • amābunt – they will love (future)
  • amā – I will love (future)
  • amāre – to love (infinitive)
  • amandus – to be loved (gerundive)
  • amā – to be loved (infinitive)

By identifying the present stem, we reduce the amount of things we need to learn. For example, why learn -āre, -ēre, -ere and -īre to form present active infinitives, when you can just learn -re and the present stem it attaches itself to?


Would that it were so simple. While this may be the underlying structure of so many verbs, things may be less clear on the surface. The addition of tense-person affixes can affect the present-stem vowel; they can shorten it, add to it, change it or delete it altogether.

For example, the addition of the present tense endings –ō, -t and -nt shortens the stem vowel if it is long. This why the present stem monē-, which has a long stem vowel, has the forms:

moneō, monēs, monet, monēmus, monētis, monent

in the present active tense. The same goes for the gerundive ending -ndus, which gives us monendus ‘to be warned’ and amandus ‘to be loved’.

Changing the stem vowel is particularly noticeable in verbs of the third conjugation. The short e that defines the group is especially susceptible to change or deletion. In the present (active) tense, it is either changed to i or u, or is absent.

legō, legis, legit, legimus, legitis, legunt

Meanwhile, in other tenses, this same e is present and even lengthened, as in the imperfect legēbam. All third-conjugation verbs follow this pattern of alteration or absence, with the exception of -iō verbs, which do not fit easily into the patterns so far discussed (for more on these, see the endnotes).

Another important example of deletion is found in the first conjugation: adding the first-person singular affix –ō removes the long ā of the stem. This gives us amō, parō, cōgitō, etc. The vowel is still present in other forms.

Lastly, the fourth conjugation, which is defined by the long vowel ī, may have its long vowel augmented with a long ē, placed between the (shortened) stem vowel and the grammatical affix. This is noticeable in the indicative imperfect tense (e.g. audiēbam) and in gerundives (e.g. audiendus)

One final use for the present stem is not to do with tenses, but with the imperative mood. This is used for giving commands and it has its own dedicated verbs forms. With the exception of irregular verbs, the second-person singular imperative (for when commanding one person) is simply the present stem!

  • amā! – ‘love!’
  • cōgitā! – ‘think!’
  • audī! – ‘hear!’
  • venī! – ‘come!’
  • pete! – ‘ask!’

To make these imperatives plural, simply add -te to the present stem (this will alter the stem vowel in third-conjugation verbs – compare pete! and petite!).

There is more to the present stem than this, but it’s enough to be getting on with. Other things behind the formation of present stems are important, but do not have such a broad import as the stem vowels outlined here. For instance, there are a few verbs that lack a vowel in the present stem. These verbs are called ‘athematic’ and include esse ‘to be’ and ferre ‘to carry’ (the infinitives of which are built from es-se and fer-re).

Moreover, there may be special affixes that have been added to the root, along with the stem vowel, to build the present stem. I’d like to discuss these in more detail in another post, but the important thing is that these affixes are often not obviously separate parts of the word. For example, the verbs tangere ‘to touch’ and crēscere ‘to grow’ have the present stem tange- and crēsce-. However, these stems really have the underlying structures tag-n-e and crēsc-e-, in which n and –sc are old affixes once productively used to build the present stem. Consequently, -n- and -sc- are absent in tenses that use a different stem, such as the perfect tense.

Speaking of which, let’s look move on to the perfect stem, the second of our three key ideas.

The Perfect Stem

The perfect stem is both fascinating and complicated. The perfect stems of Latin verbs were created in many ways, and are the result of numerous ancient and extinct processes.

The perfect stem is used to form the perfect, pluperfect, future perfect tenses and the (active) perfect infinitive. Helpfully, these verb forms all have “perfect” in their names.

To put amō, amāre into the perfect tense, we cannot use the present stem amā– and add our endings (which for the perfect are ī, istī, it, imus, istis, ērunt). Instead, we need the perfect stem, which in the case of amō, amāre is amāv.

This gives us the correct forms:

  • amāvī – ‘I loved’
  • amāvistī – ‘you loved’
  • amāvit – ‘she loved’
  • etc.

As with the present stem, there is no one rule that determines what kind of perfect stem a particular verb will have. It will at least not change for that verb.

Here are eight common ways in which Latin forms the perfect stem.

1. Adding –v– to the present stem.This is an extremely common perfect stem.
– For example, the perfect stems amāv– ‘love’ and audīv– ‘hear’ are formed through the combination of -v– and the present stems amā– and audī-.

2. Adding –u– to the bare root. This is almost the same process as the previous, since what we like to write as v and u were the same letter and basically the same sound in Classical Latin.
– For example, the perfect stems monu– ‘warn’ and habu– ‘have’ are structurally mon-u- and hab-u-. The present stems of these verbs are different, namely monē- and habē-.

3. Adding –s– to the bare root. This is also very common. It is a remnant of the aorist tense, which flourished in Ancient Greek, but died out in Latin in the shortly before the historical period – or rather, it merged with the perfect tense, which is why the Classical Latin perfect has two meanings (‘I did’ and ‘I have done’).
– The roots tēg– ‘cover’, carp– ‘pluck’ and dīc– ‘say’ have the perfect stems x-, carps– and x– (as in dīxit ‘he said’). The combination of the sounds [g] or [k] in the root and the -s- affix results in -x-.

4. Reduplicating the root. This involves taking the initial consonant of the root and ‘doubling’ it. These two consonants will be separated by a vowel, often -e-. The first consonant and the vowel together act as a kind of prefix to the root. It is not very common, but is in fact the ‘original’ way that Latin formed perfect verbs. You can find out more about reduplication here.
– The root tend– ‘stretch’ has the perfect stem tetend-. ‘You stretch’ and ‘you had stretched’ (in the present and pluperfect tenses) would therefore be tendis and tetendistī. The same goes for curre– and cucurr-, the present and perfect stems of the root curr– ‘run’.

5. Lengthening the vowel in the root. This is uncommon, but occurs in some
important verbs.
– Take Julius Caesar’s famous phrase, vēnī vīdī vīcī. The perfect stems of the roots ven- ‘come’, vid– ‘see’ and vic– ‘conquer’ are vēn-, vīd- and vīc-, with lengthened vowels.

6. The perfect stem may be the same as the root. This happens in verbs
whose roots already end in -v-/-u-, meaning that the common suffix cannot be added easily.
– For example, the root of solvō, solvere ‘loosen’ is solv-. This is also the perfect stem; ‘they solved’ (perfect tense) is therefore solvērunt.

7. Removing an affix from the present stem (and then adding another). It may not be obvious, but some present stems are in fact built around the root and an old affix. Latin removes the affix and may add the regular -v- in its place.
– As mentioned, crēscō, crēscere ‘grow’ has the present stem crēsce-. This itself is really a combination of crē– and -sc-, which adds an inchoative meaning. To form its perfect stem, -sc- is removed and -v- is used in its stead, giving us the perfect stem crēv-.

8. The perfect stem may be completely different from the present stem. This occurs in instances of suppletion – when a verb has co-opted forms from another verb and merged them into its paradigm (compare English go, went, gone). This is common in irregular verbs.
– While es- is the present stem of esse ‘to be’, the perfect stem is fu-. For example, ‘we have been’ (perfect) is fuimus and ‘to have been’ (perfect infinitive) is fuisse. Moreover, the perfect stem of ferō, ferre ‘carry’ is either tul- or reduplicated tetul-.

These eight patterns are helpfully common, but not exhaustive; many verbs exist that to a greater or lesser extent elude these rules. For example, faciō, facere ‘do’ takes its perfect stem, fēc-, from an old aorist stem, which does not easily fit into any of these patterns.

With all this variety, how on earth can we remember which perfect stem a verb has? Have I not made it all more complicated? This is where the dictionary comes in handy again.

Rule 2: To find the perfect stem of a verb, take off the ī from the first-person singular perfect active indicative form. This will be the third principle part of the verb given in a dictionary.

If you look up ‘praise’, the dictionary will give you the four forms

laudō, laudāre, laudā, laudātum

Taking off the -re from laudāre gives us the present stem laudā-. Taking off the from laudā gives us the perfect stem laudāv– (which follows the first pattern).

Unlike the present stem, the perfect stem is very robust; when we add the right tense-person affixes to it, it does not change. You can now use the perfect stem to create three tenses (perfect, pluperfect and future perfect active) and one infinitive.

For said infinitive, the affix to learn is -isse. The following are therefore perfect active infinitives, formed from the perfect stems we have already met.

  • laudāvisse – ‘to have praised’
  • tetendisse – ‘to have stretched’
  • habuisse – ‘to have had’
  • crēvisse – ‘to have grown’
  • vīdisse – ‘to have seen’
  • xisse – ‘to have covered’

Here are the same verbs in a one of the three tenses mentioned.

  • laudāvit – ‘she praised’ (perfect)
  • tetendērunt – ‘they stretched’ (perfect)
  • habuerās – ‘you had had’ (pluperfect)
  • crēverāmus – ‘we had grown’ (pluperfect)
  • vīderint – ‘they will have seen’ (future perfect)
  • tēxerō – ‘I will have covered’ (future perfect)

Some verbs may lack a perfect (and thus its associated tense-person affixes); this is the case with with deponent verbs. For example, take the common deponent verb loquor.

loquor, loquī, locūtus sum ‘speak’

Through both these forms and others, we can identify the present stem loque– and the supine stem locūt-, but no perfect stem.

The Supine Stem

The final stem type to be discussed is the supine stem. We will need this because it is used to create several important forms of ‘non-finite’ verbs (verbs that cannot stand on their own in a sentence, but are dependent on some other word). These include supines, past participles and future participles, as well as a lot of nouns that are derived from verbs. While each have their own dedicated endings, they all need this stem.

Lucky for us, there is one thing that is used to make the supine stem: the affix -t-. It is added in one of two ways.

1. Adding -t- to the present stem.
We can see this in amō, amāre, the supine stem is amāt-. We see this in the past participle amātus ‘loved’ and the future participle amātūrus ‘to be going to love’. With some verbs, the vowel of the present stem changes, as in monitus ‘warned’ from the present stem monē-.

2. Adding –t– to the root.
For example, the root of capiō, capere ‘take’ is simple cap-. This results in the supine stem capt-, as in the participle captus ‘taken’.

We observe many sound changes in the second. The last sound of the root often becomes more like the affix –t-. When two adjacent sounds become more alike, this is known as ‘assimilation’. For example, vehō, vehere ‘carry’ has the root veh-. The final -h- of veh- changes into –c- in the supine stem, which is vect-. Through this, the sound becomes a plosive sound, just like the affix -t-. Other types of assimilation are visible in many supine stems, such as scrīpt- and lēct-, the supine stems of scrībō ‘write’ and legō ‘choose, read’.

One final, slightly more complex sound change happens when the supine affix -t- is added to roots that end in –t– or –d-. The combinations, which were originally *tt and *dt, and then *ts, eventually became simply –s-.

For example, take the verbs cadō, cadere ‘fall’, vertō, vertere ‘turn’ and rīdeō, rīdēre ‘laugh’. The roots of these words are: cad-, vert-, rīd-. When -t- is added to make the supine stem, the results are cas-, vers-, rīs-.

From these we can make past participles like casus ‘fallen’, versus ‘turned’ and rīsus ‘laughed at’, from which English gets the words case, versus and risible.

How do we know the supine stem of a verb? Here’s a third helpful rule.

Rule 3: To find the supine stem of a verb, take off -um from the supine. This will be the fourth principle part of the verb given in a dictionary.

The dictionary will tell you that the verb for ‘paint’ is

pingōpingere, pīnxīpictum

The fourth form, pictum, is the supine form. To find the supine stem, take off -um. If your dictionary gives the fourth principal part as the past participle, instead of the supine, it will say pictus, so you will need remove the –us.

Now you have pict-, the supine stem of pingōpingere, you can use it to make the future participle pictūrus, the past participle pictus and nouns that build on the stem, such as pictūra ‘painting’.

An Example To Finish

So, stems. Latin verbs, like nouns and adjectives, are built with the help of stems. We should try to identify three stems for each verb – though sometimes this is not possible.

Thanks to stems, we can reduce the number of things we need to remember, because the theory can explain away irregularities and allow us to focus the affixes that all verbs share. Let’s look at one final example with two new verbs.

trahōtraheretrātractum ‘drag’
dēleō, dēlēre, dēlēvī, dēlētum ‘destroy’

By following the three rules, we can use these four forms to work out the verbs’ three stems.

  • Present stem: trahe-, dē
  • Perfect stem: trāx-, dēlēv
  • Supine stem: tract-, dēlēt

We can see from their present stems that trahe- is a third-conjugation verb, while dē– belongs to the second. As for their perfect stems, dēlēv– follows the first pattern, but trāx follows the third, with an aorist -s- added.

Now we can use these tenses to build various tenses. For example, to make the first-person plural indicative active (that is, we drag and we destroy), we need the present stem and the affix -mus.

  • trahe- + -mus –> trahimus
  • dē– + –mus –> dēlēmus

The stem vowel of trahe- is changed to -i-, as is usual. Meanwhile, the present stem remains unchanged in dēlēmus.

If we add -ndus to make two gerunds (meaning to be dragged and to be destroyed), the present stems stay strong, except for the shortened vowel of dēlē-.

  • trahe- + -ndus –> trahendus
  • dēlē– + –ndus –> dēlendus

Picking a random verb form that requires the perfect stem, we can use trāx– and dēlēv– to form third-person plural pluperfects (they had dragged and they had destroyed). The one tense-person affix we need to remember is -erant (which is an imperfect form of esse).

  • trāx– + -erant –> trāxerant
  • dēlēv– + –erant –> dēlēverant

Finally, if we want to use the past participles of these verbs (a form that is often required), we use their supine stems.

  • tract– + -us –> tractus
  • dēlēt– + –us –> dēlētus

We’ll leave it there! All in all, I hope this has successfully demonstrated what Latin verb stems are, and why they are so useful. If you have a dictionary to hand and follow the three rules for working out a verb’s stems, this system can explain a lot of irregularities and reduce the things you need to remember.


Note: -iō verbs

I have to mention these, as they do not fit comfortably in the patterns outlined. This group of verbs is so called because the first-person singular present indicative active ends in -iō, as in capiō ‘I take’. At first glance, we might think that capiō is a fourth-conjugation verb, with the present stem capī-.

However, this cannot be, as the second principle part of capiō is capere, not capīre. Moreover, the present-vowel is always short, rather than a long , as in capimus ‘we take’. The four forms are:

capiō, capere, cē, captum

The common verb faciō ‘I do’ also belongs to this group and follows a very similar pattern.

faciō, facere, fē, factum

What’s going on here? It seems to me that -iō verbs are in a sense caught between the third and fourth conjugations, their forms being a mixture of the two. Perhaps -iō verbs have two present stems: one formed with -i- (for example, capis, capiēbam) and another with -e-, which is used to build the infinitive capere and the imperative cape.

6 thoughts on “Latin Stems: Part II

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: