History and Literature | Latin Words and Grammar

What Is Vulgar Latin?

This article has been reviewed in accordance with our editorial policy.

What is Vul­gar Latin, and how does it dif­fer from Clas­si­cal Latin? As a Latin­ist or Latin enthu­si­ast, chances are that you’re going to be asked this ques­tion at some point. The answer usu­al­ly giv­en is that Vul­gar Latin was the lan­guage of the peo­ple, while Clas­si­cal Latin, com­ing down to us as a lit­er­ary lan­guage, was clos­er to how the elite spoke. This, how­ev­er, is a very simplified—and maybe not alto­geth­er accurate—picture of how things were. This arti­cle will give a fuller answer and show a few exam­ples of what char­ac­ter­ized Vul­gar Latin pro­nun­ci­a­tion, vocab­u­lary, and grammar.

But first, let me tell you a story.

You can down­load a pdf here Get a print-ready PDF ver­sion of “What is Vul­gar Latin?”.

Learning To Be Emperor

Ear­ly in 101 AD, the young Hadri­an, a favourite of Emper­or Tra­jan, had just been appoint­ed quaestor. One of his duties was to con­vey the emperor’s deci­sions to the sen­ate and recite his speech­es in his absence. With­out a doubt, Hadri­an had pre­pared well, sit­ting at his desk or roam­ing about the room read­ing the speech time and again. This was his pub­lic debut, his chance of mak­ing a good first impres­sion on a mis­trust­ful sen­ate. His­to­ria Augus­ta, a 4th-cen­tu­ry com­pi­la­tion of the lives of the emper­ors, tells us what hap­pened next:

“Quaes­tu­ram ges­sit Tra­iano quater et Artic­uleio con­sulibus; cum ora­tionem imper­a­toris in sen­atu agrestius pro­nun­tians ris­us esset, usque ad sum­mam peri­ti­am et facun­di­am Lati­nis oper­am dedit. ”

— His­to­ria Augus­ta, Vita Hadri­ani 3.1

‘He was questor in the fourth con­sul­ship of Tra­jan and Artic­uleius [101 AD]. After hav­ing been laughed at by the sen­ate while recit­ing a speech by the Emper­or in a some­what unpol­ished man­ner, he exert­ed him­self in the study of Latin let­ters to the point of the high­est pro­fi­cien­cy and fluency.’

As this account shows, Hadri­an spent much time per­fect­ing his Latin before dar­ing to open his mouth in pub­lic again. Some­how, I find this very relat­able! Of course, noth­ing is as harm­ful to author­i­ty as laugh­ter, and a future emper­or need­ed to be tak­en seri­ous­ly. Lan­guage was key to this. 


But what about Hadrian’s speech had so amused the sen­ate? He must have been one of the most cul­tured young men in the empire. Was his accent tru­ly that rus­tic? Was his ora­tion full of errors in pro­nun­ci­a­tion? Was it, dare I say, vul­gar? Though Hadri­an was like­ly born in Ital­i­ca, a Roman colony in His­pania, by the age of 14 he had already been called to Rome by Tra­jan. Antho­ny R. Bir­ley (Hadri­an: The Rest­less Emper­or, p. 46) writes that it “is hard to believe that he had picked up a ‘Span­ish’ accent in his short stay at Ital­i­ca a decade ear­li­er. Per­haps, rather, his unusu­al­ly long spell in the army and asso­ci­a­tion with cen­tu­ri­ons and rankers affect­ed his dic­tion.” Indeed, Hadri­an had spent time in Ger­ma­nia Supe­ri­or as tri­bune in the 22nd legion, which was like­ly to blame for his blun­der. Well edu­cat­ed as Hadri­an must have been, he had for­got­ten to whom he was speak­ing – if only for a moment. Per­haps he had even acquired new habits of speech. It is not rare that a per­son adapts his speech to his audi­ence, and the one that he was used to address­ing was com­posed of coarse cen­tu­ri­ons. It was in their com­pa­ny that he had spent a large part of his youth, and as we shall see, the army was a focal point of Vul­gar Latin.

By “vul­gar”, I mean it in the sense of infor­mal, col­lo­qui­al or every­day speech. Vul­gar can also sug­gest some­thing that is wide­ly spread. The Latin adjec­tive vul­garis has its ori­gin in vul­gus, ’the people/masses’, but this might be mis­lead­ing. Despite the name, Vul­gar Latin was not the lan­guage of most peo­ple. The down­trod­den slaves and day labor­ers, the orphans, beg­gars and rur­al poor have left very lit­tle in the way of lin­guis­tic evi­dence. Vul­gar Latin was the Latin of the mid­dle class. It was the Latin of peo­ple with some, but lim­it­ed, school­ing: the mer­chants, arti­sans, low­er pub­lic offi­cials and army offi­cers, who were required to know how to read and write for prac­ti­cal pur­pos­es. The mid­dle class was influ­en­tial. Army offi­cers serv­ing in every cor­ner of the empire would affect the way their troops spoke. Some­times this mil­i­tary vari­ant is called ser­mo cas­tren­sis, ‘the lan­guage of the army camp.‘ Add to this the far-trav­el­ing mer­chants, who were con­vers­ing with all sorts of peo­ple in ports, inns, and mar­kets along the trade routes. The mid­dle class was con­sid­er­ably larg­er than the refined Roman elite who could afford to spend years learn­ing how to write wit­ty epi­grams and among whom a lin­guis­tic faux pas could be the source of life­long embar­rass­ment. Sure enough, we still know of Hadrian’s blun­der after almost two millennia! 

It is in the con­text of the mid­dle class that we find the most evi­dence for every­day speech: inscrip­tions, graf­fi­ti and some­times curse tablets (defix­iones) and frag­ments of let­ters in papyrus or wood. There are also many col­lo­qui­al traits in the come­dies of Plau­tus, in the let­ters of Cicero and in lat­er Latin, often that of Chris­t­ian authors who were address­ing a wide audi­ence. Fur­ther­more, some com­pi­la­tions by gram­mar­i­ans con­tain cor­rec­tions of cer­tain words and mis­takes, but the fact is that many of these errors did not sur­vive into the Romance lan­guages that devel­oped after the fall of the West­ern Empire. Those errors may have been local or more com­mon with the less influ­en­tial low­er classes.

The Term “Vulgar Latin”

Before we begin in earnest to explore what kind of mis­takes Hadri­an could have made, it should be men­tioned that the term Vul­gar Latin is not with­out its crit­ics. The Black­well His­to­ry of the Latin Lan­guage (p. 231) states that the idea of Vul­gar Latin as a com­mon tongue for the low­er class­es is “dis­cred­it­ed among lin­guists but still tena­cious among non-spe­cial­ists.” Cer­tain­ly, we’d make a grave error if we used the term in the way that many ear­li­er schol­ars have done, posit­ing a clear dis­tinc­tion between Vul­gar and Clas­si­cal Latin, the stan­dard form of the lan­guage found in lit­er­ary works and offi­cial inscrip­tions. They were not two dif­fer­ent lan­guages and they were not iso­lat­ed from each oth­er. If this had been the case, we’d have more tes­ti­mo­ny on this from the writ­ten sources. Though clas­si­cal writ­ers some­times con­trast Latini­tas, the high­est ide­al of Latin, with ser­mo vul­garis or cot­tid­i­anus, every­day speech, this seems to be a dif­fer­ence in qual­i­ty, not nec­es­sar­i­ly in the essence of the lan­guage. The qual­i­ty or—to use a more unfash­ion­able word—the puri­ty of your Latin, as under­stood by Cicero and oth­er clas­si­cal authors, was linked to your social stand­ing. To prove your blue-blood­ed back­ground, you need­ed to speak in a cer­tain way. You’d be laughed out of the sen­ate if you mis­pro­nounced your h’s!

The Black­well His­to­ry is right that Vul­gar Latin was not uni­form, but not even Clas­si­cal Latin—as a lit­er­ary language—was as rigid as is some­times believed. Take as an exam­ple one of the most basic fea­tures of Latin gram­mar, the sequence of tens­es. No one clings to it as firm­ly as Cae­sar, not even Cicero. Livy has his own traits, often pre­fer­ring pri­ma­ry tens­es in indi­rect speech where oth­ers would have used sec­ondary tens­es. Con­sid­er­ing this and many oth­er vari­a­tions in what are regard­ed as the paragons of Gold­en Age prose, the spo­ken lan­guage must have been even less stan­dard­ized. In fact, this is part of what is typ­i­cal for Vul­gar Latin, in the sense of col­lo­qui­al, non-elite, Latin: it was not fixed in the way that Clas­si­cal Latin was, whether in spelling, gram­mar or vocab­u­lary. There must have been more vari­a­tion between the dif­fer­ent provinces and cities of the empire, between the dif­fer­ent social class­es and pro­fes­sions, and even between the old­er and younger generations.

So why is it still a use­ful term? 

The answer is that the Romance lan­guages of today have many traits which are rare or com­plete­ly miss­ing in Clas­si­cal Latin lit­er­a­ture but are com­mon for what has been called Vul­gar Latin. Though ear­li­er def­i­n­i­tions of Vul­gar Latin may have been flawed, per­haps the solu­tion is not to stop using the term, but to be more pre­cise in what we mean by it. Vul­gar Latin was a soci­olect (or group of soci­olects) main­ly of the mid­dle class from the time of Plau­tus to a cen­tu­ry or so after the fall of the West­ern Empire. It was less stan­dard­ized in its gram­mar and vocab­u­lary than Clas­si­cal Latin, but many of its dis­tinc­tive traits were com­mon through­out the Latin-speak­ing part of the empire. The rea­son for this was that its speak­ers were high­ly mobile and influ­en­tial. Many oth­er schol­ars, includ­ing the authors of The Black­well His­to­ry of the Latin Lan­guage, pre­fer the some­what wider term sub-elite Latin. This is worth com­mit­ting to mem­o­ry, but for the pur­pos­es of this arti­cle, I will con­tin­ue refer­ring to the lan­guage as Vul­gar Latin, see­ing as this term is in wider use.

How Was Vulgar Latin Pronounced?

Before we delve deep­er into any vul­gar traits, one ques­tion is espe­cial­ly impor­tant to con­sid­er: How can we know how Latin was pro­nounced? Besides the Latin alpha­bet itself, which was cre­at­ed on the basis of the spo­ken lan­guage, the pri­ma­ry sources are lit­er­ary ref­er­ences, loan­words and spelling mis­takes (in graf­fi­ti and inscrip­tions) as well as the mod­ern Romance lan­guages. Cicero him­self dis­cuss­es pro­nun­ci­a­tion in many of his pri­vate let­ters. Oth­er writ­ers, such as Aulus Gel­lius (2nd cen­tu­ry AD), can also give valu­able infor­ma­tion, as can lat­er gram­mar­i­ans com­plain­ing about com­mon mis­takes. A work of some inter­est for our inquiry into Vul­gar Latin is Appen­dix Pro­bi (3rd or 4th cen­tu­ry), which is found in the same man­u­script as the gram­mat­i­cal work Insti­tu­ta artium, ascribed to a cer­tain Probus (pos­si­bly Mar­cus Valerius Probus). Appen­dix Pro­bi con­tains a list of com­mon mis­takes in vocab­u­lary and pro­nun­ci­a­tion togeth­er with the cor­rect forms. 

As stat­ed above, loan­words from Latin into oth­er ancient lan­guages, main­ly Greek, but also Goth­ic, are anoth­er source that lin­guists can use. For instance, the Greek Καῖσαρ and Goth­ic 𐌺𐌰𐌹𐍃𐌰𐍂 (Kaisar) show that Latin Cae­sar must have been pro­nounced with a hard /k/ in antiquity. 

Inscrip­tions in stone are also an impor­tant source, but since they are so for­mu­la­ic, clues to pro­nun­ci­a­tion are most­ly found in uncon­scious mistakes—what Leonard Palmer calls “the occa­sion­al inad­ver­ten­cies.” As such, they often con­tain one or only a few errors, but can oth­er­wise be with­out fault. The stone carvers who made these inscrip­tions are the per­fect exam­ple of peo­ple who had lim­it­ed edu­ca­tion in the Roman world. Like the mer­chants and offi­cers, their use of the writ­ten lan­guage was for prac­ti­cal rea­sons only, not for high literature.

To pick up where we left off: What kind of vul­gar pro­nun­ci­a­tion did Hadri­an acquire dur­ing his time in the Roman army? If it was an into­na­tion or a rhythm that seemed rough to the learned ear, then it is almost impos­si­ble to say what it sound­ed like. How­ev­er, there are many traits, com­mon in his army camp, that could have occurred in his speech.

One that comes to mind con­cerns the diph­thong /ae/. There was a ten­den­cy already dur­ing the late repub­lic to pro­nounce it as a monoph­thong, prob­a­bly as [ɛː], if it stood in an unac­cent­ed posi­tion. This devel­op­ment con­tin­ued dur­ing the 1st cen­tu­ry AD, spread­ing to the pro­nun­ci­a­tion of /ae/ in accent­ed posi­tions. It is unsure to what extent the cul­tured elite tried to retain the diph­thong. They cer­tain­ly did so in writ­ing, and it would not be sur­pris­ing if they had made an effort in the spo­ken lan­guage as well. Nev­er­the­less, inscrip­tions reveal the con­fu­sion of non-elite Latin speak­ers. Hyper­cor­rec­tions (i.e. mis­tak­en cor­rec­tions) like “baene” for “bene” (‘well’), though rare, are evi­dence of this. Some also show enclitic -que (‘and’) spelled like “quae”, as in the funer­ary mon­u­ment of Petro­n­ia Hedo­ne, now locat­ed in the Muse­um of Fine Arts, Boston. 





‘Petro­n­ia Hedo­ne made this for her­self and for Lucius Petro­n­ius, son of Phile­mon, and for her freed­men and freed­women and their descendants.’

Funerary inscription in stone from early 2nd century with busts of a Roman woman and a young boy.
Funer­ary inscrip­tion, ear­ly 2nd. cen­tu­ry A‑D- CIL VI.24037. Muse­um of Fine Arts, Boston, pho­to Vic­tor Frans.

Anoth­er exam­ple that I hap­pened to see in the Gre­go­ri­ano Pro­fano Muse­um in the Vat­i­can last Autumn had “aeo­rum” for “eorum” (‘their’, CIL, VI, 2365 & 2366), of which there are sev­er­al oth­er exam­ples from around the time of Hadri­an (AE 1983, 0086, AE 1984, 0129, CIL III, 01808). A num­ber of curse tablets (defix­iones) found in Roman Africa, in what is today Tunisia, dis­re­gard the diph­thong in “dae­mon”, sev­er­al of them begin­ning with the words “adi­uro te demon qui­cunque es…” (‘I entreat you, spir­it, who­ev­er you are…’; Audol­lent, Defix. Tab. 265, 286, 290, 291, 293–295).

Despite the sound change hav­ing tak­en place many cen­turies before, the digraph spelling <ae> remained in use for a long time in offi­cial inscrip­tions. But by the 5th cen­tu­ry AD, even these show an <e>. A case in point is the fol­low­ing tile from the reign of the Goth­ic king Theoder­ic the Great (ruled 493–526 AD), who tried to show his good will to the peo­ple of Rome by restor­ing pub­lic works. 

Roman tile stamped with +REG(NANTE) D(OMINO) N(OSTRO) THEODE + RCO BONO ROME.
Tile stamped with +REG(NANTE) D(OMINO) N(OSTRO) THEODE + RCO BONO ROME. Case Romane del Celio, Rome. Pho­to: Vic­tor Frans.

This tile has the stamp: “Reg­nante domi­no nos­tro Theoder[i]co, bono Rome”, ‘In the reign of our lord Theoder­ic, for the good of Rome.’ “Rome” would nat­u­ral­ly be “Romae” (dativus com­mo­di) in Clas­si­cal Latin!

When it came to oth­er vow­els, long /eː/ and short /i/ coa­lesced into a sound some­where in between the two, usu­al­ly writ­ten [ẹ] in pho­net­ic nota­tion. In some instances this is evi­denced by the con­fu­sion of <i> and <e>, as in the fol­low­ing inscrip­tion on a 1st cen­tu­ry AD sepul­chre (CIL VI, 35337) in the Capi­to­line Museums:







‘To the Manes. For Fruc­tus, the ser­vant of Domitian’s wife Domi­tia. He lived six years and twen­ty days. His moth­er made it for her well-deserv­ing son and for her­self and her kins­men and their descendants.’

This exam­ple shows the carver’s uncer­tain­ty in how to repro­duce [ẹ], which led to the spelling of ficit for fecit. This is quite an ear­ly exam­ple, but the trait is recur­ring all through­out Roman impe­r­i­al his­to­ry. A lat­er 4th cen­tu­ry AD inscrip­tion fea­tures the word tri­bunus (‘tri­bune’) spelled as <tre­bunus> (AE 1907, 0143).

A trait which most cer­tain­ly exist­ed among Hadrian’s sol­diers was the drop­ping of ini­tial /h/. Already in the late repub­lic it was pro­nounced “leniter et leviter” (‘gen­tly and soft­ly’, from the poem below) by the elite, but cer­tain peo­ple were inclined to drop it alto­geth­er. A few tried to over­com­pen­sate for this ten­den­cy, as is evi­denced by Cat­ul­lus ridi­cul­ing a cer­tain Arrius for putting an /h/ where it didn’t belong, say­ing “hin­sidias” (‘ham­bush’) for “insidias” (‘ambush’), pro­nounc­ing it with the same vig­or as his rus­tic forebears:

Cat­ul­li car­men LXXXIV ad Arrium

Chom­mo­da dice­bat, si quan­do com­mo­da vel­let 
dicere, et insidias Arrius hin­sidias,
et tum mir­i­fice sper­abat se esse locu­tum,
cum quan­tum poter­at dix­er­at hin­sidias.
Cre­do, sic mater, sic Liber avun­cu­lus eius,
sic mater­nus avus dix­er­at atque avia.
Hoc mis­so in Syr­i­am requier­ant omnibus aures;
audibant eadem haec leniter et leviter,
nec sibi pos­til­la metue­bant talia ver­ba,
cum subito affer­tur nun­tius hor­ri­bilis:
Ion­ios fluc­tus, postquam illuc Arrius isset,
iam non Ion­ios esse sed Hio­n­ios.

‘Arrius said “ghains”, if he ever want­ed to say “gains”, and “ham­bush” for “ambush”, and hoped he had spo­ken admirably well, hav­ing said “ham­bush” as vig­or­ous­ly as pos­si­ble. The same way, I think, did his moth­er, his uncle Liber [or ‘freed­man uncle’] and his mater­nal grand­fa­ther and grand­moth­er speak. When he was sent to Syr­ia, our ears could final­ly rest! They heard the same words spo­ken gen­tly and soft­ly, and did not fear them, when sud­den­ly a ter­ri­ble mes­sage was brought: The Ion­ian waves, after Arrius had gone there, were no longer “Ion­ian”, but “Hion­ian!”

Painting called "Catullus" with Romans reading outside under the trees, by Stefan Bakalowicz.
Ste­fan Bakalow­icz, Cat­ul­lus (1885)

Though the elite might have tried to main­tain ini­tial /h/, by Hadrian’s time it had all but dis­ap­peared from com­mon speech. There are clues to this devel­op­ment in curse tablets (defix­iones), that drop <h> in “hoc” and “hac” (‘this’), as in: “ex oc die, ex ac ora” (‘from this day, from this time’, Audol­lent, Defix. Tab. 229) and in: “ex oc die” (Audol­lent, Defix. Tab. 287).

The let­ter <b>, when stand­ing between two vow­els (and some­times after a vow­el and a nasal /m/ or /n/), was from the 1st cen­tu­ry AD onwards increas­ing­ly pro­nounced like a voiced bil­abi­al frica­tive [β]. You can hear how [β] and oth­er pho­net­ic sym­bols are pro­nounced on this web page . Since this sound is some­where between /b/ and /w/ (the con­so­nan­tal pro­nun­ci­a­tion of <u>), con­fu­sion arose in some speak­ers as to how these let­ters should be spelled. Appen­dix Pro­bi, a 3rd or 4th cen­tu­ry list of cor­rec­tions of mis­takes in Latin, points to the mis­pro­nun­ci­a­tion or mis­pelling of the word ‘com­mon peo­ple’: “plebes non ple­vis.” The exam­ple of [β] in the 3rd cen­tu­ry mosa­ic below, from the ruins of a fishmonger’s shop in Ostia Anti­ca, might seem puz­zling. A pos­si­ble rea­son (giv­en by Gio­van­ni Becat­ti, a promi­nent schol­ar and exca­va­tor of Ostia) why cus­tomers were asked to step on a dol­phin, is that dol­phins were believed to scare away all the fish. Anoth­er the­o­ry has it that is is instead the octo­pus in the dolphin’s mouth that was meant as the object of the visitor’s scorn. What­ev­er the truth might be, the mosa­ic works as an exam­ple of an apotropa­ic image, a ward against harm or bad luck. It fea­tures <b> in “inbide”, which should be “invide” (‘envi­ous one’).

Black and white mosaic of fish from Ostia Antica with the text "INBIDE, CALCO TE".
Mosa­ic in Ostia Anti­ca, 3rd Cen­tu­ry AD. “INBIDE, CALCO TE.” ‘Envi­ous one, I step on you!’ Pho­to: Vic­tor Frans.

An oppo­site exam­ple occurs in this 2nd cen­tu­ry funer­ary inscrip­tion from the necrop­o­lis at Por­ta Medi­ana, Cumae. Here, /b/ spelled as <v> can be seen in the words “fem­i­nae incon­par­avili” (‘to an incom­pa­ra­ble woman’). 

Funer­ary inscrip­tion, Por­ta Medi­ana, Cumae (2nd cen­tu­ry ad). Pho­to: Vic­tor Frans.

Besides this con­fu­sion regard­ing inter­vo­cal­ic /b/, there was also a lat­er ten­den­cy dur­ing the 2nd cen­tu­ry AD, par­tic­u­lar­ly in Italy, to pro­nounce ini­tial /b/ or /w/ as a voiced bil­abi­al frica­tive [β] (a sound which, as men­tioned, lies between these two), but this pro­nun­ci­a­tion seems to have dis­ap­peared before it was passed on to the Romance lan­guages. The cur­rent pro­nun­ci­a­tion of some Span­ish speak­ers of ini­tial /b/ as an approx­i­mant of [β] hap­pened much later.

Anoth­er vul­gar trait that does not seem to have left many traces was pro­nounc­ing the diph­thong /au/ as /oː/. Indeed, Cicero’s adver­sary Claudius Pul­cher once ensured that he’d be renamed Clodius after being adopt­ed into a Ple­beian fam­i­ly! That the same trait was still alive in the 1st cen­tu­ry AD can be glimpsed from an enter­tain­ing anec­dote from Sue­to­nius about the emper­or Ves­pasian: “Mestri­um Flo­rum con­sularem, admoni­tus ab eo ’plaus­tra’ potius quam ’plostra’ dicen­da, pos­tero die ’Flau­rum’ salu­tavit.” ‘After the ex-con­sul Mestrius Florus had remind­ed him to say “plaus­tra” instead of “plostra” (‘wag­ons’), the fol­low­ing day he greet­ed him with the name “Flau­rus”’ (De vita Cae­sarum, De Ves­pasiano XXII.I). How­ev­er, the devel­op­ment of /au/ to /oː/ seems to not have been passed on to the Romance lan­guages. Though there is an /o/ in French or and in Span­ish and Ital­ian oro (‘gold’) for Latin aurum, this change hap­pened much lat­er. In fact, Roman­ian and Occ­i­tan still have aur, and sim­i­lar forms can be found in var­i­ous dialects in the West.

There are many oth­er vul­gar traits in pro­nun­ci­a­tion worth explor­ing, but in this arti­cle, I’ll just give a quick run­down of the most impor­tant ones:

  • Final /m/, as in noun end­ings like -um or -em and in verbs forms such as -eam or -am, was very light­ly pro­nounced even dur­ing the late repub­lic. It grad­u­al­ly began dis­ap­pear­ing from com­mon speech. A 1st cen­tu­ry AD exam­ple of this can be seen in the Vat­i­can Muse­ums (in Gre­go­ri­ano Pro­fano) on a funer­ary inscrip­tion made “in honore[m] L(ucii) Caci Rebur­ri f(ilii)” (‘to the hon­or of Lucius Cacus, son of Rebur­rus’), where “in hon­ore” is miss­ing a final <m> (CIL XIV, 00413 (1)). One result of the loss of final /m/ was a nascent con­fu­sion between the accusative and ablative.
  • /x/ and /ps/ was often assim­i­lat­ed into a dou­ble /ss/. This is already clear in Pom­peian graf­fi­ti, where we find “isse” for “ipse” (‘he him­self’). Like­wise, in lat­er inscrip­tions from oth­er parts of Italy, it’s not rare to see “vis­it”, or even “visse”, for “vix­it” (‘he/she lived’). True enough, in mod­ern Ital­ian the same form is “visse.”
  • A lat­er trait, which was like­ly not yet com­mon in Hadrian’s day, was that short /u/ was grad­u­al­ly approach­ing an o-sound. Latin cum (‘with’) is spelled as <con> (as lat­er in Ital­ian and Span­ish) in a let­ter writ­ten on papyrus by a sol­dier from the 2nd cen­tu­ry AD (CEL 146). How­ev­er, the change seems to not have tak­en place in some parts of the empire, for exam­ple in Sar­dinia and Dacia. Exam­ples are found in Appen­dix Pro­bi, giv­ing the cor­rec­tion of the word for ‘girl’ as “puel­la non poella.”

We’ll get to the gram­mar soon enough, but this is a good place to note that sound change was one of the rea­sons why the cas­es start­ed to dis­ap­pear, even­tu­al­ly becom­ing only two (nom­i­na­tive and accusative/ablative) in the West (lat­er to be lost), and three in Dacia (nominative/accusative, dative/genitive, vocative).

Vulgar Latin Vocabulary

We’ve delved into the pro­nun­ci­a­tion of Hadrian’s times, but there are more aspects of Vul­gar Latin that are impor­tant for under­stand­ing its devel­op­ment. For instance: How was Vul­gar Latin vocab­u­lary dif­fer­ent from that of Clas­si­cal Latin?

A fea­ture that can be traced back to repub­li­can Latin is the use of diminu­tives in col­lo­qui­al speech. Diminu­tive end­ings make a word more famil­iar, inti­mate or in some cas­es con­de­scend­ing, like “Britun­culi”, ‘those lit­tle Britons’, an expres­sion found in let­ters from Roman sol­diers in Vin­dolan­da. Diminu­tives were cre­at­ed with a suf­fix, adding -culus, -cula or -culum (depend­ing on gram­mat­i­cal gen­der) or sim­i­lar forms to the end of a word. Plau­tus (Asi­nar­ia 668) has “pre­hende auri­culis” (‘grab [me] by [my] lit­tle ears’). Diminu­tives are also found in the let­ters of Cicero, main­ly when he is writ­ing to close friends. In these we find words like lit­teru­lae, ‘a lit­tle let­ter’ and febric­u­la, ‘a lit­tle fever.’ Ad Heren­ni­um, a 1st cen­tu­ry BC guide to rhetoric, gives a sam­ple show­ing how to speak in a rhetor­i­cal­ly low­er style (adten­u­a­tum genus). There are many col­lo­qui­alisms here, but most notice­able is the word oric­u­lae, ’lit­tle ears.’ Much lat­er, in the 3rd or 4th cen­tu­ry AD, Appen­dix Pro­bi warns against the same exact word (in its syn­co­pat­ed form): “auris non ori­cla” (‘ear, not lit­tle ear’). In fact, many of the cor­rec­tions in Appen­dix Pro­bi seem to indi­cate that diminu­tives were tak­ing the place of the orig­i­nal word, and true enough, Romance lan­guages often only have the diminu­tive form left. Vul­gar Latin auri­cla (giv­en as oric­u­la/ori­cla in the exam­ples above) became orec­chia (lat­er most­ly orec­chio) in Ital­ian, ore­ja in Span­ish and oreille in French.

Incised wall-plas­ter from Her­cu­la­neum (CIL IV, 10520), ca 1–79 AD. Pho­to: Vic­tor Frans.

The graf­fi­ti above is an exam­ple of diminu­tives (and oth­er vul­gar traits) from Her­cu­la­neum. It’s not easy to make out the text, but cat­a­logues give it as “con­si­tonts Her­cu­lane­ses navcu­lae”, which in Clas­si­cal Latin would read “con­sis­tunt Her­cu­la­nens­es nav­ic­u­lae” (‘The Her­cu­lanean ships make a stop’). 

Some oth­er diminu­tives from Appen­dix Pro­bi:

  • “nep­tis non nep­ti­cla” (‘grand­daugh­ter, not lit­tle granddaughter’)
  • “anus non anu­cla” (‘old woman, not lit­tle old woman’)
  • “fax non facla” (‘torch, not lit­tle torch’)

As can be seen in “ori­cla” from “oric­u­la”, syn­cope (the loss of a sound inside the word, also present in the exam­ple of the graf­fi­ti from Her­cu­la­neum) was often remarked upon in Appen­dix Pro­bi. The gram­mar­i­an gives more exam­ples of this, as in “frigi­da non fric­da”, ‘cold’ (refer­ring to “aqua frigi­da”, ‘cold water’), which has some clear coun­ter­parts in Ital­ian fred­do/fred­da, Span­ish frío/fría and froid/froide in French. It is also impor­tant to remem­ber that Romance lan­guages could bor­row Latin words in lat­er times, as in Span­ish frígido/frígida from frigidus/frigida, which there­fore do not show the same sound changes.

In many cas­es, where Latin had two fair­ly close syn­onyms of the same con­cept, only one sur­vived into the Romance lan­guages. This is most often the one which had been com­mon in col­lo­qui­al speech. A few exam­ples are list­ed below: 

  • The adjec­tive bel­lus increas­ing­ly replaced pul­cher (‘beau­ti­ful’). It had exist­ed before, but with a slight­ly dif­fer­ent nuance, clos­er to ‘pret­ty, fine’, not rarely in a sar­cas­tic sense, as in Cicero (Facete dic­ta 21.1): “O hominem bel­lum!” (‘Oh, what a fine fellow!’)
White marble bust of Cicero against a red background.
Bust of Mar­cus Tul­lius Cicero, Capi­to­line Muse­ums, Rome. Phot: Vic­tor Frans.
  • Oppidum (‘town’) and urbs (‘city’) were often replaced by civ­i­tas, which orig­i­nal­ly meant ‘state’, as in Sp. ciu­dad, It. cit­tà, Fr. cité.
  • Cabal­lus (‘horse, pack horse’) was nor­mal­ly not used in Clas­si­cal Latin prose, but occa­sion­al­ly in its poet­ry. It might orig­i­nal­ly have been a Celtic loan­word. Grad­u­al­ly, how­ev­er, it replaced the usu­al equ­us (‘horse’), even­tu­al­ly result­ing in Span­ish cabal­lo, Ital­ian cav­al­lo, French cheval and Roman­ian cal.

Some words changed mean­ing and replaced oth­er words, such as homo (‘human being, per­son’), which took the place of vir (‘man’), as is lat­er evi­denced by Ital­ian uomo, French homme, Span­ish hom­bre and Por­tuguese homem.

Some new and inge­nious cre­ations were also made in Vul­gar Latin: While Clas­si­cal Latin would have advenire (‘arrive’), which also sur­vives in many Romance lan­guages, the Vul­gar Latin arri­pare, hav­ing its ori­gin in ad ripam (‘to shore’), also has a few descen­dants, like Ital­ian arrivare, Span­ish arrib­ar and French arriv­er.

Vulgar Latin Grammar

As we have seen, the cas­es were slow­ly dis­ap­pear­ing in the devel­op­ment of Vul­gar Latin. They were often replaced by prepo­si­tion­al phras­es or sim­ply dis­ap­peared (like the vocative).

  • Gen­i­tive reg­is (‘of the king’) was often sub­sti­tut­ed by “de rege.”
  • Dative regi (‘for/to the king’) could be giv­en as “ad rege(m).” 
  • Pure abla­tive, that is when the abla­tive was used with­out prepo­si­tions, for exam­ple in some sep­a­r­a­tive con­texts, like domo (‘from home’), or in the instru­men­tal glad­io (‘with the sword’), was increas­ing­ly replaced by prepo­si­tion­al phras­es, like “a/de domo” or “cum gladio.”
  • Voca­tive was grad­u­al­ly lost except in set phras­es, for exam­ple domine (‘lord’) in Chris­t­ian con­texts. Even in the works of ear­ly writ­ers like Plau­tus you can find the nom­i­na­tive instead of the voca­tive, as in “meus ocel­lus” for “mi ocelle” (‘oh, my sweet­heart’, Asi­nar­ia 664), but some of these exam­ples might be due to the meter. Though lost in the West, voca­tive is still present in Romanian.

There are more devel­op­ments in Vul­gar Latin gram­mar wor­thy of note.

Charles H. Grand­gent calls Vul­gar Latin “more explic­it” than Clas­si­cal Latin. One thing that today’s stu­dents of Latin often strug­gle with is the lack of arti­cles: fem­i­na might mean ‘woman’, ‘a woman’ or ‘the woman.’ Appar­ent­ly, the same dif­fi­cul­ty faced the Romans of antiq­ui­ty. Because of this, in their dai­ly speech they some­times used unus/una/unum (‘one’) as an indef­i­nite arti­cle, like Eng­lish a/an before a noun. There are even exam­ples of this in Plau­tus: “una ader­it muli­er” (‘a/one women will be present’, Pseudo­lus 948) and “unus ser­vos vio­len­tis­simus” (‘a/one very impetu­ous ser­vant’, Tru­cu­len­tus 250). Like­wise, per­son­al pro­nouns had orig­i­nal­ly been used only for empha­sis, since the verb forms often gave enough clues. In Vul­gar Latin, pro­nouns like ego and tu slow­ly lost their empha­sis and start­ed being used more fre­quent­ly. Like­wise, ille and illa took on the roles of def­i­nite arti­cles (as Eng­lish ‘the’), result­ing in Ital­ian il and la, French le and la and Span­ish el and la.

Gram­mat­i­cal gen­der was some­times con­fused. Petro­n­ius (mid 1st cen­tu­ry AD) – though he might have been par­o­dy­ing the bro­ken Latin of Greek freed­men – has some of his char­ac­ters say vinus and caelus for vinum and caelum (‘wine’ and ‘sky’), mis­takes pos­si­bly attrib­uted to Greek οἶνος and οὐρανός. How­ev­er, con­fu­sion of gram­mat­i­cal gen­der can also be attest­ed among native speak­ers, espe­cial­ly regard­ing mas­cu­line and neuter, which were lat­er to merge. Also, neuter plur­al forms, like cas­tra (‘camp’) were often inter­pret­ed as fem­i­nine sin­gu­lar. This can be seen in Appen­dix Pro­bi, which makes the fol­low­ing cor­rec­tion: “vico cas­tro­rum non vico cas­trae” (‘street of the camp’).

Verbs were not spared the new developments.

Fre­quen­ta­tive verbs were more com­mon in col­lo­qui­al speech. Even­tu­al­ly, many of these start­ed to take the place of the verbs from which they orig­i­nat­ed. Latin canere was replaced by cantare, ‘to sing’ (Fr. chanter, It. cantare, Sp. can­tar) and like­wise iacere by iactare/iectare, ‘to throw’ (Fr. jeter, It. gettare, Sp. echar).

In the per­fect pas­sive, as in “mis­sus est” (‘he was sent’), the “est” (and cor­re­spond­ing forms) was more often replaced by the per­fect form “fuit” etc. You can occa­sion­al­ly see this even in clas­si­cal writ­ers (as in Sal­lust and Valerius Max­imus), but it is more preva­lent in Late Antiq­ui­ty. Even His­to­ria Augus­ta (4th cen­tu­ry AD), which most­ly tries to give the cor­rect forms, at one point has “usus fuer­at” for “usus erat” (‘he had used’). This con­struc­tion became the norm in the Mid­dle Ages, which, admit­ted­ly, makes these exam­ples some­what uncer­tain, since they might be changes made by medieval scribes. 

The mod­ern Romance lan­guages most­ly cre­ate their per­fect forms by using the verb for ‘to have.’ Thus, “I have seen him” would in Span­ish be ”lo he vis­to”, in Ital­ian “l’ho vis­to” and in French “je l’ai vu.” The ori­gin of this form can already be seen in Plau­tus, as in “insti­tu­ta … cunc­ta … defor­ma­ta habebam” (‘I had all my plans arranged’, Pseudo­lus 677). This exam­ple demon­strates that using habere with an objec­tive pred­ica­tive was only a small step away from the per­fect. Com­pare also the expres­sion “com­per­tum habere” (‘to know full well, to have ascer­tained’) found in clas­si­cal authors.

Old wall painting from Pompeii from before 79 A.D. showing Romans gambling, etc.
Wall Paint­ing from the Caupona of Salvius, Pom­peii (before 79 A.D.; CIL IV 3494) Pre­suhn, Emil. 1882. Pom­pe­ji. Die neuesten aus­grabun­gen von 1874 bis 1881. Abtheilung v, pl. 36. The paint­ing is now in the archae­o­log­i­cal muse­um of Naples.

One last thing: Do you remem­ber the pic­ture above the intro­duc­tion? It might not be easy to see, but the words above their heads give an enter­tain­ing lit­tle exchange between two gamblers!

Man 1: “Exsi.” (Cor­rect­ed: “Exii.” ‘I went out!’ i.e. ‘I won!’)

Man 2: “Non tria, duas est.” (Cor­rect­ed: “Non trias, dyas est.” ‘Not a three, it’s a two.’)

Man 1: “Noxsi! A me tria. Eco fui!” (Cor­rect­ed: “Noxi! A me trias. Ego fui!” ‘You crook! A three by me! I was (the winner)!’)

Man 2: “Orte, fel­la­tor! Eco fui!” (Cor­rect­ed: “Orte, fel­la­tor! Ego fui!” ‘Ortus, you c********r! I was (the winner)!’)

Innkeep­er: “Itis foras, rixsatis!” (Cor­rect­ed: “Ite foras, [ibi/foris] rix­am­i­ni” ‘Go out­side and quarrel!’)

Today, the wall paint­ing is dam­aged to such a degree that we must rely on old­er draw­ings like the one above. There are a few uncer­tain­ties in the dia­logue: Is “Orte” a name, like Ortus or Hor­tus (in Pom­peii the <h> is some­times left out), or, less like­ly, a par­tici­ple (mean­ing ‘born’)? Should the num­bers be inter­pret­ed as trias and dyas, a ‘tri­ad’ and a ‘dyad’, orig­i­nal­ly Greek loan­words? There are also some pecu­liar­i­ties of Pom­peian graf­fi­ti on dis­play here, such as eco for ego and <x> spelled like <xs>. One trait regard­ing gram­mar is the dis­ap­pear­ance of the depo­nent verbs (that have a pas­sive form but an active mean­ing); “rixsatis” is an exam­ple of this, where the imper­a­tive “rix­am­i­ni” (‘quar­rel [out­side]!’) is like­ly intend­ed. Alter­na­tive­ly, it could be the indica­tive with the same form (‘you’re quar­relling’) or the sub­junc­tive “rix­em­i­ni” (‘you may quar­rel [out­side]’).

Toward The Romance Languages

The gap between lit­er­ary and non-elite Latin widened dur­ing Late Antiq­ui­ty. Cer­tain­ly, sound changes that start­ed out as vul­gar were often adopt­ed by the elite lat­er on, but the sit­u­a­tion was slight­ly dif­fer­ent for gram­mar and vocab­u­lary. Read­ing some of the best 4th cen­tu­ry AD writ­ers, one could hard­ly guess how much had hap­pened in every­day speech from the days of Cicero. Still, it was like­ly that the schools in the cities and small towns of the West had a con­serv­ing effect on the lan­guage of many Latin speak­ers. The elites who could afford a thor­ough edu­ca­tion, and the mid­dle class­es who went to school only to learn the basics, prob­a­bly had the declen­sions and con­ju­ga­tions ham­mered into them. Gram­mar­i­ans waged a con­stant war on pleonas­tic expres­sions, faulty syn­tax, sole­cisms, and bar­barisms, hold­ing the most vul­gar nov­el­ties at bay. How­ev­er, once the West­ern Empire col­lapsed, so did most of the school­ing sys­tem. After not too long, it was main­ly church­men who helped their flocks and young ini­ti­ates to pre­serve a mod­icum of cor­rect speech. 

But when was Latin no longer Latin? When did it become the pre­de­ces­sors of the Romance lan­guages? An easy (but per­haps arbi­trary) way of deter­min­ing when the lan­guage should no longer be called Latin, would be to ask when the dif­fer­ent tongues of the Latin-speak­ing world stopped being mutu­al­ly intel­li­gi­ble. When would a Spaniard no longer under­stand a Dacian? It might be impos­si­ble to say, since this peri­od has much less evi­dence for every­day speech. Anoth­er way of deter­min­ing a bound­ary would be when Latin writ­ings could not be under­stood by the com­mon peo­ple. In rela­tion to this, there is evi­dence that saints’ lives in Latin were read in church­es and under­stood by con­gre­ga­tions in Gaul as late as the 7th cen­tu­ry. But already a few gen­er­a­tions after this there are ref­er­ences to lin­gua Romana (‘the Roman lan­guage’) as dis­tinct from the Latin of the church. Should we there­fore say that the 7th or 8th cen­tu­ry forms the end of Latin as a native language?


Vari­a­tion in every­day speech has always been so great that any bound­ary in time or descrip­tion we give of it—nay, even the def­i­n­i­tions themselves—will with­out doubt be sim­pli­fi­ca­tions. Nev­er­the­less, they can help us in under­stand­ing the broad­er strokes of the lan­guage. Vul­gar Latin gives us insight both into what it is not, i.e. Clas­si­cal Latin, and into what it would become, the Romance lan­guages of today: Por­tuguese, Span­ish, French, Ital­ian, Roman­ian and more. It also paints a pic­ture of oth­er denizens of Rome and its empire than we nor­mal­ly meet in clas­si­cal lit­er­a­ture. I’m not the only one, I’m sure, who has dreamt of hear­ing the Latin of the forum, of the mar­kets and tav­erns of ancient Rome. Yet the thought of a moth­er remem­ber­ing her six-year-old son in a sim­ple Latin inscrip­tion or of Hadri­an grap­pling with the speech of high soci­ety gives us an inkling that Latin once was a liv­ing lan­guage for every­one, not just for scholars.


The Vul­gar Latin traits that I’ve addressed here are just a selec­tion. Assured­ly, a whole book would be need­ed to describe all the dif­fer­ences between Vul­gar and Clas­si­cal Latin and the devel­op­ment of the Romance lan­guages. Luck­i­ly enough, there has been much writ­ten on these sub­jects. An acces­si­ble cat­a­logue of Vul­gar Latin traits is found in Grand­gent (1907), reprint­ed in 2009. Leonard Palmer (1954) pro­vides an absorb­ing and elo­quent account of the his­to­ry of the Latin lan­guage. A more updat­ed work is The Black­well His­to­ry of the Latin Lan­guage (2007), which con­tains a chap­ter on sub-elite Latin. The dis­cus­sion there is main­ly based on three let­ters – one by a mer­chant and the oth­er by two sol­diers – which are a fas­ci­nat­ing read. 

To study a larg­er selec­tion of texts, I’d rec­om­mend An Anthol­o­gy of Infor­mal Latin, 200 BC–AD 900 (2016), edit­ed by J. N. Adams, one of the lead­ing experts on sub-elite Latin. The data­base which has been most help­ful in find­ing inscrip­tions is the Epi­graph­ic Data­base Roma. The quotes from Plau­tus most­ly come from the Exem­pla Plauti­na in Rohlfs (1969), but some have been found by search­ing PHI Latin Texts.

Adams, James N. (ed.). 2016. An Anthol­o­gy of Infor­mal Latin, 200 BC–AD 900: Fifty Texts with Trans­la­tions and Lin­guis­tic Com­men­tary. Cam­bridge, UK: Cam­bridge Uni­ver­si­ty Press.

AE = L’An­née Épigraphique

Audol­lent, Defix. Tab. = Defix­ion­um tabel­lae quotquot innotuerunt. Audol­lent (ed.)

CEL = Cor­pus Epis­tu­larum Lati­narum Papyris Tab­u­lis Ostracis Servatarum

CIL = Cor­pus Inscrip­tion­um Latinarum

Clack­son, James & Hor­rocks, Geof­frey C. 2007. The Black­well His­to­ry of the Latin Lan­guage. Malden, MA: Blackwell.

Grand­gent, Charles H. 2009 [1907]. An Intro­duc­tion to Vul­gar Latin. Rich­mond: Tiger Xenophon.

Her­man, József. 2000. Vul­gar Latin. Trans­la­tion by Roger Wright. Penn­syl­va­nia State Uni­ver­si­ty Press.

Palmer, Leonard R. 2001 [1954]. The Latin Lan­guage. Lon­don: Bris­tol Clas­si­cal Press.

Pre­suhn, Emil. 1882. Pom­pe­ji. Die neuesten Aus­grabun­gen von 1874 bis 1881.Leipzig: T.O. Weigel.

Rohlfs, Ger­hard. 1969. Ser­mo vul­garis Lat­i­nus: Vul­gär­lateinis­ches Lese­buch. 3rd edi­tion. Tübin­gen: Niemeyer.

Victor Frans

Victor Frans

Victor Frans is currently writing his dissertation on Saxo Grammaticus. He holds a BA in Latin from Stockholm University and an MA in Medieval Studies from the University of Oslo. He has worked in the Swedish National Archives and on a project on St. Birgitta of Sweden (14th century).
Written by Victor Frans

Written by Victor Frans

Related articles

The Latin of Saxo Grammaticus

The Latin of Saxo Grammaticus

The Hunt for Saxo  Shortly after the year 1208, Saxo Grammaticus put the finishing touches to his great ...
The Supine in Latin Grammar: What it is and What its Function is

The Supine in Latin Grammar: What it is and What its Function is

Among Latin’s many verb forms, the supine, causes students quite a lot of confusion. In this article, I will ...