How are painting and picture related English words? Why does Latin vincō ‘I win’ become vīcī in the past tense? How did the same Latin verb give English both convince and conviction?
This blog post is about the wonderful world of the nasal infix, a linguistic phenomenon that comes in very handy for people learning Latin. It discusses its history, its function and where and how to spot it in Latin and some other languages.
It’s also intended as (probably) the final chapter in a series of posts on the irregularities of Latin verbs – which together I hope will demonstrate how reasonable the language’s verbs actually are, or will at least convey a general impression of ‘don’t panic!’ You can find the previous part here.
Confuse/confound and picture/painting are pairs of English words. In terms of their origins, the pairs are each etymological doublets – two separate words in the same language with the same origin. Specifically, confuse/confound and picture/painting come down to us from two Latin verbs, the first verb meaning ‘to mix together’ and the second ‘to paint’.
In both pairs, we can see that one word has something that the other lacks: namely, the nasal consonant /n/. It’s there in confound and painting, yet absent from the roots of their sister words confuse and picture. This is not something that has simply been lost over time during the words’ journeys into English; it’s a difference that goes all the way back to their Latin ancestors.
For example, confound comes from the Latin verb cōnfundere, specifically forms of the verb that are built on the present stem cōnfunde-. (A stem is like a midpoint between the fundamental root of a word and its completed, usable form – for a recap to what we mean by “stem”, follow this link.)
Meanwhile, confuse comes from the past participle cōnfūsus, which is instead built on a different verb stem – the supine stem cōnfūs-. This is still the same Latin verb with the same meaning, but it is two different stems of that verb that have led to English confuse and confound.
Likewise, painting, with its nasal /n/ sound in the root, goes back the present stem pinge-, while picture derives from pict-, the nasal-less supine stem of the same verb. Once again, the present stem includes -n-, which will therefore be found in all verb forms in Latin that use the present stem, such as the present infinitive pingere ‘to paint’ or the present-tense pingit ‘she is painting’.
We’ve hit upon something key here! The -n- in pingere is more than a meaningless sound or letter. It’s actually an addition to the root, something that, through its presence, has helped to turn the fundamental root pic- into the present stem pinge-. In the same way, the present stem cōnfund- can be analysed as formed of the prefix cōn-, the root fud- and an inserted -n-.
Things which are added to roots to supplement their meanings are known as affixes. Since -n- is added to the inside of the root, it is an infix. Being a nasal consonant, it’s known as the nasal infix.
Thanks to Latin, it’s made its way into lots of English words, such as comprehend, tangent and infringe. Knowing it’s there means you know you can take it out – which in turn helps you to make new etymological connections!
But how does it work?
Many verbs in Latin include the nasal infix and across its appearances, we see that its association with the present stem is very strong. To better appreciate this, let’s take a look at some verbs for ‘win’, ‘break’ and ‘burst’.
We can say that the fundamental roots of these verbs, which are what express these three meanings, are vic-, frag- and rup-. Out of these roots, Latin makes three types of stem for its verbs to use: present, perfect and supine.
To make their supine stems, Latin simply adds -t- to the root. This results in vict-, frāct- and rupt-. It is from these supine stems that English gets victory, conviction (but not convince), fraction, fracture, rupture, corrupt, abrupt and many more words.
For the perfect stem, the vowel in the root changes: vic-, frag– and rup– become vīc-, frēg– and rūp-. Since the perfect stem is used to form the perfect tense, the stem vīc– is there in vīcī ‘I conquered’, as in Julius Caesar’s most famous phrase.
What matters most to us is the present stem. To make it, these verbs show the addition of the nasal infix to the inside of the root. This results in vince-, frange- and rumpe-. The third shows assimilation between adjacent sounds, the infix becoming labial –m– before labial –p-. The resulting present stem is used to form not only present-tense verbs (e.g vincit ‘she wins’) but also imperfect verbs (e.g. rumpēbant ‘they were bursting’), future verbs (e.g. frangam ‘I will break’) and present infinitives (e.g. vincere ‘to win’).
When it comes to these three verbs, these are the stems we need to learn:
- Present stem: vince-, frange-, rumpe-
- Perfect stem: vīc-, frēg-, rūp–
- Supine stem: vict-, frāct-, rupt–
As mentioned, the nasal infix is only present in the present stem. This gives us a big clue as to its original purpose.
With all this information in mind, we can better explain the four principal parts that the dictionary will give us for these verbs, and the variation between them, because the four principal parts intentionally make use of all three stems.
|First P. P.|
|Second P. P.|
|Third P. P.|
|Fourth P. P.|
I have won
I have broken
I have burst
If only that were the end of the story! Latin, being a living language, did somewhat muddle this system through the completely understandable process of ‘levelling’. That is to say, Latin levelled the differences between the stems of some verbs, adding to the nasal infix to where it did not originally belong. This very sensibly made different forms of the same verb resemble one another more closely and it suggests to us that the infix had long since lost any obvious grammatical functions.
One example of this is the already mentioned verb pingere ‘to paint’. Its supine stem may be pict-, but its perfect stem is pīnx-, which is made up of the root pic-, the common aorist affix -s- and additionally the nasal infix -n-. This means that the present stem pinge– and the perfect stem pīnx- have been made more similar. Other verbs that show this levelling between the present and perfect stems are fingere ‘to form’ and pandere ‘to spread out’.
- Present stem: pinge-, finge-, pande-
- Perfect stem: pīnx-, fīnx-, pand-
- Supine stem: pict-, fict-, pass–
There are even verbs in which the nasal infix has ended up in all three stems. Pāns– is a possible alternative supine stem for pandere, while iungere ‘to join’, from which English gets join and junction, consistently shows this thorough levelling.
- Present stem: iunge-
- Perfect stem: iūnx–
- Supine stem: iūnct-
Although ubiquitous, we believe that the -n- of these stems is nonetheless an addition, because of related words that lack it, such as iugum ‘yoke’, from which English gets jugular.
The nasal infix is very old; it must go all the way back to Proto-Indo-European, because we find it in Indo-European languages other than Latin. For example, it’s also a handy tool for learning Ancient Greek. It explains why we find present-tense verbs like lambánō ‘I take’, which in the aorist tense is the un-nasal élabon.
It is much more noticeable in Sanskrit. One of its verb classes, the seventh, is even traditionally defined by the presence of the nasal infix, including verbs like yunákti ‘to join’ (related to yoga). It is found in other classes too, such as the sixth-class verbs muñcáti ‘to free’ and lumpáti ‘to rob’, all of which lack the nasal when conjugated in different tenses.
One thing that remains constant across these languages is the infix’s association with the present tense – or, to be precise, with the imperfective aspect. This conveys the idea that the action of the verb is incomplete and ongoing, something that verb forms with the nasal infix have in common. A link between ongoing actions and the present time seems natural and this explains its later continued association with the present stem and tense.
The infix was therefore one of the ways in which Proto-Indo-European could construct imperfective-aspect verbs out of basic roots. Infixes are comparatively rarer than prefixes and suffixes (especially among Indo-European languages), and so it has been proposed that the nasal infix started life as an suffix, following the root, before being metathesized to a position within it.
As the nasal infix lost productivity as a morpheme and fizzled out over time, its use became less and less systematic. There are numerous cases which, with etymological analysis, show us that the infix sometimes ended up distinguishing not just two separate stems, but two whole verbs with separate meanings. The Latin verb pandere, mentioned previously, is a sister word of the verb patēre ‘to open’ – seemingly two distinct verbs by the time of the Romans, but actually born of one root and disconnected by the presence of the infix.
A natural question to ask is whether the nasal infix can be found in other branches of the Indo-European family – perhaps in English? The answer to the first is yes; it is by no means limited to the three usual suspects of Indo-European studies. As for English – well, it’s possible. One verb has been proposed (and subsequently disputed) as containing the old infix: stand, a present-tense form, which in the past tense changes to stood.
All in all, the nasal infix is not something Latin learners need to lose sleep over. It is not something that will affect your translations, especially if you are translating from Latin into your own language.
It is, however, something that can explain a lot of irregularity among forms of the same verb and I believe that’s never a bad thing. It is, I like to think, quite fun to spot and can (probably) offer you some fun English etymology facts for when the pubs are open again.