An Invaluable Tool for a Latinist’s Toolkit
If there is one thing that I urge all Latin learners to get to grips with, it is the concept of stems. My passion for this topic is so great that I have structured my own Latin course around it, including and using it from the outset. I owe my own fluency in the language to learning about stems back when I was first learning Latin, and now I am witnessing its explanatory power again with my own students.
Like so many things in linguistics, the idea of stems has the ability to demystify a lot of what is irregular and unpredictable in ancient languages like Latin, Greek and Sanskrit – things that have put many people off these languages altogether. This article, which I fully intend to be helpful for all Latin enthusiasts, from beginners to experts, will focus on stems in Latin, though I think much of it will be applicable to other languages too.
Stems can really help you to learn Latin, I swear!
What is a stem? And why haven’t I heard of it before?
The concept of linguistic stems is normally very new for Latin learners, in part because most of us aren’t used to thinking about linguistic ideas, but also because it is not a useful concept for the study of the English language.
Latin constructed its nouns, adjectives and verbs by means of stems. When I say ‘construct’, I am referring to how words are built out of component parts. For example, the English word disinfected is constituted of three evident parts: dis-, infect and -ed. Though two of these are not stand-alone words, English speakers will still have an intuition about their meaning and how to use them.
The process of building words in English is simpler than it is in Latin. It involves only roots, affixes and complete, ready-to-use words. The root of a word is what gives the word its meaning. It is the meaningful ‘core’ of the word. In disinfected, the root is infect. If you’re historically inclined, you could analyse this too into in– and -fect, but looking at English as it is today, infect is a single unit with a complete meaning.
In English, roots can be stand-alone words; nothing else is required. For example, the words
- beer
- happy
- red
- sing
are all roots and complete words. If we add some affixes to these roots, the resulting words are also stand-alone words, but have become more complicated than bare roots.
- beers (suffix -s makes the root plural)
- unhappy (prefix un- makes the root negative)
- redden (suffix -en makes the root into a verb)
- singing (suffix -ing makes the root into a participle)
However, in Latin, the formation of words has more steps. When it comes to nouns, adjectives and verbs, roots are only rarely allowed to be complete words, without any additions. Instead, words are formed through a three-step process.
(1) root –> (2) stem –> (3) complete word
How does this look in terms of the actual component parts of the word? Most (but not all) Latin nouns, adjectives and verbs have the following structure:
root + stem affix + grammatical affix
This is already getting rather technical. For now, let’s simply say that in nearly all Latin nouns, adjectives and verbs, we should be able to identify three parts: (1) the meaningful root (at or near the beginning), (2) some grammatical stuff at the very end and (3) an extra something sandwiched between the two. We can see this in action with everyone’s favourite delight, the Latin noun.
Part I: Latin Nouns (and Adjectives!)
In nouns and adjectives, stem affixes are a single vowel. They are essentially meaningless; they contribute nothing to either the meaning or the grammar of the completed word. On one side of the stem affix is the root, while on the other is what I call the ‘case-number affix‘. This is the grammatical part of the word, which tells us what case it is in, whether it is singular or plural, and, for adjectives, what gender it has.
We need an example. Here are six Latin words, all nouns, all in the accusative case and singular, but a mixture of masculine and feminine genders.
| fēminam | ‘woman’ | campum | ‘field’ |
| ducem | ‘leader’ | turrim | ‘tower’ |
| frūctum | ‘fruit’ | diem | ‘day’ |
What is the one thing that they all have in common? It’s the final consonant, -m. It is this single thing that expresses the accusative case and singular number. The rest of the word is the stem – that is, the combination of the root and a meaningless stem vowel.
Chopping up nouns into three parts leaves us with this, with the stem vowels in bold:
| fēmin – a – m | camp – o – m |
| duc – m | turr – i – m |
| frūct – u – m | di – ē – m |
Now, there are some surprises here, but first, let’s appreciate what the idea of stems can do.
Traditionally, without the concept, we would have to learn four things for (masculine and feminine) accusative singular nouns, namely, -am, -um, -em and -im, as well as when to use them. This is because stem vowels and case-number suffixes are usually lumped together under the ambiguous label of ‘endings’. Given that the traditional endings multiply case-number affixes by stem vowels, there are many to learn. With the help of stems, we only have to learn one case-number affix (-m), along with the unchanging stem of the noun that it is attached to.
The addition of word-final affixes can have an affect on the original stem. We see this in the examples above. It’s said that the stem of campum is campo-, because it seems that –o– is the ‘underlying’ stem vowel, which appears in some contexts, but is altered in others.
| Singular | Plural | |
| Nominative | campus | campī |
| Genitive | campī | campōrum |
| Dative | campō | campīs |
| Accusative | campum | campōs |
| Ablative | campō | campīs |
| Vocative | campe | campī |
Moreover, if we go back to Old Latin, we find the forms we expect with the -o- vowel. For example, we find words like campos for classical campus, and campoī for campī. If we hop across to Ancient Greek, the -o- of the second declension is still present and strong. I should also mention that the vocative -e in o-stem nouns is not something added to the stem, but is a change to the stem vowel that goes back much further that Latin.
As for the analysis of ducem as duc-m, this happens in nouns that have no stem vowel; we say that these words have ‘consonant stems’. In ducem, the root and stem are the same (duc-) and the -e- is present between the stem and the affix -m only to aid pronunciation. We call this an epenthetic vowel.
Why is the -e- in ducem not also a stem affix? This is because in other forms of dux, there is no vowel between the stem duc- and the grammatical affix. For example, the nominative singular dux is formed from only duc– and -s.
Through the lens of stems, we can see how this nominative singular affix (-s) on dux is exactly the same one as we see at the end of many other words, like dominus, frūctus and fidēs (deconstructed: domin-o-s, frūct-u-s and fid-ē–s).
So, based on the stem suffix, we can identify six stem categories. Every Latin noun and adjective belongs to one category (or sometimes is a mixture of two – see below).
| a-stems (e.g. fēmina, mensa, nauta) | o-stems (e.g. dominus, donum) |
| consonant stems (e.g. dux, rēx, tempus) | i-stems (e.g. turris, mare, sedīle) |
| u-stems (e.g. senātus, frūctus, genū) | ē–stems (e.g. spēs, fidēs) |
These six categories match up fairly well to the traditional five declensions.
- a-stems – first declension
- o-stems – second declension
- consonant stems – third declension
- i-stems – third declension
- u-stems – fourth declension
- ē-stems – fifth declension
The third declension of Latin nouns is made up of both consonant stems and i-stems, since both types of stems attach themselves to the genitive singular affix -is. It’s the genitive singular that is the defining characteristic for assigning a Latin noun to a declension. Understanding this can demystify much of the third declension, which is, in my opinion, the hardest to learn.
During the classical era, Latin was going through the natural process of changing i-stem words into consonant stems. The Classical Latin that we are taught captures this change as if frozen in time. Some i-stem words stay strong to their original structure, maintaining the stem vowel -i- between the root and the case-number affixes, and so are said to have ‘pure’ i-stems.
For example, feminine turris ‘tower’ and neuter mare ‘sea’ are in the nominative singular built of turr-i-s and mar-i. Additionally, because mare ends in -e, the stem vowel -i, without a final case-number affix to shield it, lowers to -e. This is part of a regular sound change that be seen in other neuter words like sedīle and omne. In other forms of these words when the stem vowel isn’t the final sound, -i- can still be found. For instance, the -i- is present in maria and marium, two plural forms of mare.
Meanwhile, some third-declension nouns have a mixture of i-stems and consonant-stem endings. For example, urbs ‘city’ shows no -i- in the nominative singular, nor in the forms urbem or urbe, but it is there in the genitive plural urbium. Other words in this category are nox ‘night’ and nūbēs ‘cloud’.
The third declension can seem bewildering, but so much of the confusion is due to this single process, the reassignment of stems.
In light of all of this, I have also tried to create, to help Latin learners, a set of what the case-number affixes of Latin really are – that is to say, when they are separated from stem vowels. It is not easy and not perfect, as the fusion of the two parts can make their origins unclear. Regardless, I believe this will be useful. If anything, it may calm fears about the size of the task of learning Latin nouns, because it is certainly much smaller than the traditional tables for noun endings!
Here’s how to read it:
- It plots gender and number against case.
- For example, matching “Neut. Pl” (neuter plural) against with the accusative case gives the affix a. Indeed, all neuter accusative plural nouns end in -a.
- Multiple affixes in one cell mean that more than one affix is possible for nouns of that gender and number.
- For example, “-, e” means that the ablative singular is shown by either e or a long stem vowel (e.g. urbe or turrī, dominō, diē).
- ∅ stands for no obvious affix added.
- We see this in nominative singular fēmina, in which -a is just the stem vowel.
- It is also very common in singular neuter nouns in the nominative and accusative. For example, genū, mare, cornū, tempus, cor and caput all have no affix added. The exception to this are o-stem neuters, which take the affix –m in the nominative and accusative singular (e.g. bellum, donum, initium).
- A hyphen in this table stands for a long stem vowel (or epenthetic vowel).
- In the case of -s, this means that a long vowel precedes the affix, as in dominōs or fēminās.
- Affixes in brackets are what I consider simply variants of the form preceding the brackets.
- For example, I believe that -rum is a variant of um, used for the first and second declensions and perhaps formed when um was added to the accusative plural of the noun.
- The simple affix i refers to the vowel [i], regardless of length.
All that remains is for you to combine the affix with the stem of the noun.
How they interact can look strange, but the processes are regular and learnable. Adding affix to stem may affect the stem in various ways. For example, some case-number affixes completely supplant the stem vowel.
- fēmina– + īs (dat./abl. pl.) –> fēminīs
- frūctu– + ibus (dat./abl. pl.) –> frūctibus (frūctubus is indeed also possible)
- domino– + ī (nom. sg.) –> dominī
- bello– + a (nom. pl.) –> bella
Some affixes just modify the final sound of the stem. This is noticeable in consonant stems (third declension nouns). Take the consonant stem rēg- ‘king’.
- rēg- + s (nom.sg.) –> rēx
- rēg– + m (acc.sg.) –> rēgem
- rēg– + ibus (dat./abl.pl.) –> rēgibus
Or the consonant stem virtūt- ‘manliness, virtue’
- virtūt– + s (nom.sg.) –> virtūs (Latin did not like the sound combination ts and routinely simplified it to just s)
- virtūt– + e (abl. sg.) –> virtūte
- virtūt– + um (gen. pl.) –> virtūtum
The combination of an a-stem with the affix i creates the vowel -ae. It makes sense if you pronounce the two together as a Roman would. Take for example fēminae ‘women’, made of the stem fēmina– and the suffix i. This is why I’ve put in the table that the genitive singular affix for both the first and second declensions is just i!
- domino– + i (gen.sg.) –> dominī ‘of the the master’
- fēmina– + i (gen.sg.) –> fēminae ‘of the woman’
- bello– + i (gen.sg.) –> bellī ‘of the war’
One final thing to mention – a bonus bit of grammar! All of this can be applied to Latin adjectives. Traditionally, the practice is to divide adjectives into two camps: those that take first-/second-declension endings, and those that take third-declension endings. This is a good analysis, which stems only serve to complement.
The adjective bonus ‘good’ is a first-/second-declension adjective. In its endings we can we both o-stems and a-stems used to express all three genders.
Meanwhile, third-declension adjectives are more fixed; each adjective belongs to one of three familiar groups: i-stem, consonant stem or mixed. Here are examples of each group:
- vetus ‘old’ is a consonant-stem adjective. The appropriate case-number-gender affix is added directly to vetus, which, with rhotacism, results in forms like veteris and veterum.
- omnis ‘all’ is an i-stem adjective. The -i- is present in nearly all forms, such as omnia and omnium. However, in the neuter nominative singular (in the absence of a case-number affix), the vowel lowers and we get omne. Note that also in this group are common adjectives that end in -ālis, such as mortālis ‘mortal’.
- audāx is a mixed-stem adjective. It is caught between a consonant stem audāc- (hence its forms audāx, audācem) and the i-stem audāci- (hence audācia, audācium).
So, that’s it for the moment! In Part II, we’ll look at verbs and how their stems work. Though slightly different to their noun and adjective counterparts, stems are just as important and, I hope, just as useful.
Danny L. Bate 
3 thoughts on “Latin Stems!”