Wanana sculon Francon

a linguist´s look into Carolingian everyday life

Introduction

The Parisian Conversations are probably the most enigmatic piece of Old High German literature that ever came down to us. What we have preserved in them is an actual and unique picture of the Carolingian day-to-day. Vernacular renderings of conversations in which people greet each other, laugh together and insult each other. The Parisian Conversation may have originated in an Early Medieval travelguide of sorts intended for Romance speaking clergymen traveling to the Franconian speaking east. It is almost as if we can see the people who uttered these sentences a thousand years ago standing on some withered Roman road, the cobble stones green with moss for the lack of maintenance, a hot wind blowing through the trees and the reed roofs of large farmhouses laying hazily in the background. A historian’s dream come true!

Strangely, however, the importance of the Parisian conversations is mostly overlooked by historians of Carolingian culture (even nowadays). Their linguistic and philological value on the other hand have been acknowledged by Old Germanicists from the nineteenth century onwards. The glosses were written down by a Romance speaking clerc, who learned Old High German as a second language and possibly heard the sentences somewhere during his travels. The "Latin" sentences that the Old High German was supposed to gloss can hardly be called Latin anymore, and is better understood as colloquial Proto-Romance. A good reminder to the historian that the clergymen writing our manuscripts did not "live" in a Latin world, they lived in a vernacular world and spoke to eachother using the Romance or the Germanic vernacular.

The Old High German sentences are concatenations of spoken language heavily influenced by the way the Romance speaking scribe pronounced Old High German, naturally with Romance dialect features. The copyist who subsequently copied the text into the manuscript that we nowadays posses probably didn’t know any Old High German and corrupted parts of the text by making transcription errors and wrong word divisions. But still, the fact that we have an actual specimen of original non-poetic spoken Old High German, which is in no way influenced by Latin examples, makes the Parisian conversations a very valuable text for linguists.

 But not only the specialized linguist or philologist will enjoy the content of the Parisian Conversations. The vulgar and bawdy nature of quite some parts of the text make for a good laugh and a fun read. It seems very likely that some conversations were only written down for their humorous nature in the first place. In this article I want to present you a scandalous collection of sentences from the Parisian Conversations, which may together constitute a somewhat coherent converstation. I will give you the Old High German gloss, a tentative Old High German reconstruction of what the author of the text actually wanted to transcribe and the Latin translation. Finally I will give a modern rendering of the conversation as it might have looked. I will, however, not give a linguistic analysis, for that would make a whole article on itself.

Manuscript

The Parisian Conversations are preserved in the margins of a ninth century manuscript[1] that originated in the south of France, possibly the monastery st. Marcel at Châlons-sur-Sâone. The Parisian Conversations are copied from an older source that might be connected to the area of Sens, possibly the area where the text originated. The text is written in a dialect of Old High German which has some peculiar idiosyncracies. This has led some scholars to believe that it actually represents a dialect of Franconian spoken in the north of France, an elite Frankish sociolect surrounded by Romance speakers. Most of the idiosyncracies however can be explained as orthographical alternations and pronunciation difficulties common to the Romance speaking scribe. Please note that the Romance-Latin glosses are just as interesting to the linguist as the Old High German. 

Text

I place the last two sentences which were numbered 83 and 84 in Wilhelm Braune’s edition in the same context as the conversation. I do this because sentences 83 and 84 were written on the upper margin of the manuscript page and the text seems to continue in the right margin of the page, therefore 83 and 84 seem to be followed by sentence 60 in Braune’s way of counting. The fact that 83-84 and 60-67 share a rather scandalous nature combined with the possibility that they may have followed eachother up in the original text, brings me to put the two sets of sentences together, for that may be how the text was originally intended.

                I interpret the lines as representing a verbal jousting amongst friends. The inquiry as to why one of the speakers wasn’t present at matins does not sound like something you would ask a stranger. The use of the politeness form in the second plural is used as part of their verbal game.[11] The rudeness of the retort seems to confirm this interpretation. Sentence 66 is best interpreted as referring to a physical beating by a riding whip. Sentences 83 and 84 use the word sarden which is cognate to Old Icelandic serða, which meant “unconventional intercourse”. The Latin translation fottire is the Romance continuation of Classical Latin futuere and seems to have a quite general meaning.

Reconstructed conversation

In my reconstruction of the text I will name the speakers Ruodlieb and Walthar, after the protagonists of the two most important secular poems that were written down in Carolingian times.

Ruodlieb:              How many times did you have sex?

Walthar:               Truly, I dont know!

Ruodlieb:              Why weren’t you there at matins?

Walthar:               I did not want to go!

Ruodlieb:              Did you sleep with a woman in your bed?

                               If your lord finds out that you slept, by my head, he will be so angry!

Walthar:             What did you say?

Ruodlieb:              Listen, fool!

Walthar:               Do you want to feel the skin of your horse on your back!?!

Ruodlieb:              Fool, he likes having sex (too much)!

Bibliography

Wolfgang Haubrichs et Max Pfister eds., „In Fracia Fui; Studien zu den romanisch-germanischen Interferenzen und zur Grundsprache der althochdeutschen ‚Pariser (Altdeutschen) Gespräche“, Abhandlungen der geistes- und sozialwissenschaftlichen Klasse 1989, 6 (Mainz 1989).

Braune, Wilhelm, Althochdeutsches Lesebuch (17th edition 1994 Tübbingen; 1875).

On the divide between history and historical linguistics concerning the Migration Age with the Scandinavian origin of the Goths as case study.

Sometimes neighbouring academic disciplines do not speak in the same idiom. Nowhere is this more clear than in the controversial subject of the “transformation of the Roman world”, a line of inquiry into the dynamics of the transition periode between late antiquity and the early Middle Ages favoured by leading historians such as Walter Goffart, Ian Wood and Peter Heather. It’s argument is summarized by Robert Anderson, director of the British Museum:

The period of transition between the world of late antiquity and the Middle Ages has [...] traditonally been seen as one of chaos and obscurity, the “Dark Ages”. Yet modern scholarship is increasingly revealing how profoundly dynamic and influential were the cultural and intellectual shifts which mark the period. Far from initiating an age of barbarism, the successor states saw themselves as part of a Roman continuum, and readily exploited the institutions and intellectual traditions of late antiquity, adapting and reinventing them to suit their own changing circumstances and cultural traditions. (Anderson 1997; 8)

Revisionism of this pivotal period in Western History was not new. Since the end of the second world war scholars became mighty uncomfortable discussing the period in terms of “Germanic expansion”, the nazi discourse of “Germanic fraternity” freshly in mind. The “transformation of the Roman world” movement could be seen as an exponent of this postwar “uncomfort” and has become increasingly influential in “correcting” popular views of late antiquity and the early Middle Ages. A challenge to modern revisionism came from Bryan Ward-Perkins in 2005. Whereas the revisionists may sometimes suggest an easy and peaceful integration of Germanic peoples into a continuing and evolving Roman world, Ward-Perkins takes fault with this line of interpretation and says he is “convinced that the coming of the Germanic peoples was very unpleasant for the Roman population, and that the long-term effects of the dissolution of the empire were dramatic” (Ward-Perkins 2005: 10). His book “The fall of the Roman empire and the end of civilization” is an impressive polemic for rehabilitation of those late antique sources that speak of catastrophe, massive invasion and crumbling city walls. This is in sharp contrast to historians who dispute that potent barbarians groups marched into the Roman empire (e.g. Noble 2006; xvi) and those that argue hardly any migrations at all had taken place (Bowersock 2000:187-97).

                How does this relate to scholars working in the field of Germanic studies, who used to draw information from a wide interdisciplinary spectrum of philology, historical linguistics and folklore? For one, the divide between the disciplines has widened. Han Nijdam (2001) characterizes the revisionism of historians and folklorists after the second world war in the following way:

“suddenly there were no relics of Germanic customs or ideas anymore, neither in the Middle Ages nor in the Early Modern Period, but everything was Christian from the moment missionaries build churches and monasteries. Literary sources were often written by them and for every costum or idea a fitting bible passage could be found.” (Nijdam 2001; 60)

This assertion is even more true for the revisionism by contemporary historians. Whereas scholars in Germanic studies assume that along with the survival of the Old Germanic languages also other cultural products survived that could be named Germanic, historians who work in the “transformation of the Roman world” paradigm reject such an assertion[1]. For them Germanic identity arose anew in the margins of the limes (i.e. the Roman border) and was barely inherited from prehistoric times. For them Germanic identity arose from the cultural dialectic between Roman rural society and small groups of barbarian immigrants.

For every comparative Indo-European linguist the inadequacies of such a paradigm are evident. The parallels in literary motives, metre and even specific formulas between Celtic, Greek, Vedic, Slavic and Germanic cosmogenic writings are astonishing. They are hardly explicable by any other means than assuming heritage from a common prehistoric literary reservoir, a reservoir justly called Indo-European (e.g. see Puhvel Comparative Mythology 1989). The comparative linguist may draw the same conclusions for the non-christian / non-Roman parallels in Old Germanic literary monuments e.g. the vestiges of a Common-Germanic pantheon (such as in the case of the worship of Wodan and Thunar). Common heritage from a cultural tradition which, in analogy to our linguistic terminology, may be called Proto- or Common-Germanic, seems likely. Rejections of such a theoretical construct by Noble and Goffart (2006: 12) are unjustified for they ignore empirical data mined from the field of comparative mythology and comparative law arguing in favour of it.

But to indicate the divide between historians and comparative linguists in the area of migration history I’d like to discuss the case of the Gothic migration, which is basically the case of the Scandinavian origin myth. This myth is to be found in Jordanes[2]’ Getica, an abridgment of an earlier work, historia gothorum, by the Gothic historian Cassiodorus recounting the history of the Goths. Jordanes wrote it at the Byzantine court when the Ostrogothic kingdom of Italy had fallen under renewed Byzantine attempts of reconquest. According to this myth, Scandinavia was the womb of peoples from which the Goths set out in three ships. From the Baltic they made their way to the Black Sea where they arrive in the scope of Roman historians. The veracity of this myth stands at the center of the problem.

The main historical polemic on this topic is between Herwig Wolfram and Walter Goffart. Herwig Wolfram is a student of Reinhard Wenskus on whose work on ethnogenesis he largely builds. Wolfram formulated a thesis later known as the Wenskus-Wolfram thesis that focusses on the leading role in ethnogenesis for so called “nuclei of tradtion” (Traditionskerne). These are ancient families whose connections to the past gave a common focus for the diverse ethnic groups within a multi-ethnic confederation. The multiethnic groups would associate and identify themselves with these ancient families and accept their tribal name (Wolfram 2006: 52-54). His argument is that, although Gothic identity has been reinterpreted and adapted multiple times throughout Gothic history, Jordanes is basically right in recounting the Scandinavian origin myth. This is how the tribal name of the 2^nd century Gutones in Tacitus’ Germania survived in the late antique Γότθοι of Procopius. Goffart rejects Wolfram’s arguments as reading truth in mere fiction and argues for a strict 6^th century Byzantine creation ex nihilo. That this cannot be true is made clear by the research of Svennung (1967: 235) into the ethnonyms cited by Jordanes, which seem to be authentic correspondences to Scandinavian tribal names.

In the field of comparative Indo-European linguistics a Scandinavian origin for the late antique Goths is widely accepted. Augustin Speyer (2007) states:

“[Das Ostgermanische ist] Ursprünglich im südlichen und südöstlichen Skandinavien beheimatet; die Inselnamen Gotland und Bornholm (< Borgundarholmr) weisen auf Beziehung dieser Plätze zu ostgermanischen Stämmen hin.”

Jasanoff (2008) is more careful in placing the original home of the Goths in Scandinavia, but the connection to the Baltic Gutones is not in doubt. He states this view as follows:

“Like other East Germanic tribes such as the Vandals, Burgundians, Gepids and Heruls, the Goths originally lived in the area of present-day Poland and eastern Germany; their own traditions place their earliest homes in southern Sweden.”

An interesting but very late source contiguous to this Scandinavian origin myth is known to most Old Germanicists but remains undiscussed by most historians (Wenskus on the other hand treats it in some detail). This is the Gutasaga, composed around 1220 in Scandinavian Gotland but preserved in a fourteenth century manuscript, which tells us that a third of the people left Gotland in ships and traveled to the Baltic coasts, from whence they traveled through Eastern Europe to arive in the land of the Greeks.

Sīþan af þissum þrim aucaþis fulc ī Gutlandi sō mikit um langan tīma, at land elpti þaim ai alla fyþa. Þā lutaþu þair bort af landi huert þriþia þiauþ, sō alt sculdu þair aiga oc miþ sīr bort hafa, sum þair ufan iorþar āttu.

Afterwards from these three, the people on Gotland multiplied throughout such a long time that the land could not feed them all. Then they drew lots so that a third of the people left the land, so that they would have and take with them everything which they possessed above the earth.

The Gutnic myth runs parallel to Jordanes account in some respects and deviates from it in others. Historical scepticism towards a continuity of an oral traditional preserving the migration of the Goths from Gotland is however justified.  From a linguistic perspective there is also some evidence to substantiate Jordanes’ claim of a Scandinavian origin. The following parallels between North-Germanic and Gothic have led some scholars to argue for a Gotonordic prestage (Eckhardt Meinecke 1953: 83-84).

1.      Pgmc -i̯i̯-> Goth -ddj- and Pgmc -i̯̯i̯-> OIce. -ggj-

Pgmc -ṷṷ- > Goth -ggw- and Pgmc -ṷṷ- > OIce. -ggv-

2.  retainment of Pgmc *-z# as Gothic -s# and OIce. -r#

3.      retainment of the fourth inchoative class of weak verbs in Gmc -nan

4.      lexical parallels

5.      Loss of the verbs preserved in Wgmc dōn, gān and stān.

6.      retainment of 2.sg.pret. ending of the strong verbs in *-t, e.g. Goth and OIce. gaft against OHG gabi

However striking the parallels, they could also be explained in terms of retaining archaic features than in terms of common innovation. West-Germanic and North-Germanic share more features than an alleged Gotonordic. The inconclusiveness of the linguistic evidence does not give us an argument in favour of the Scandinavian origin myth of Jordanes. Nonetheless,  a linguistic continuity from the 2^nd century tribal name Gutones to the fourth and fifth century Gothi and Γότθοι seems probable.

Also to be mentioned is the archaism of the Gothic language itself. As a Germanic language attested in fourth century writings it predates the literary monuments of most other Old Germanic languages in atleast three and a half hundred years, so some archaism is to be expected. Most “transformation of the Roman world” historians on the other hand would have the fourth century Goths live in a highly Romanized ethnically diverse cultural realm and considering this historical background the lack of cultural loans from Latin and Greek is perplexing (not counting eclessiatical terminology). This might point to a strong Germanic core to Gothic identity (contra Goffart), an assumption strengthened by the fact that Gothic supplied the neigbouring Alans with quite some lexical material which eventually made it’s way into modern Ossetic, while the other way around the loans seem to be quite limited. Abaev (Moscow 1958) mentions, amongst others, these words:

                Gothic                                   Ossetic

Goth aggwus               > Ossetic. wyngæg “narrow, oppressed, surpressed.

Goth qairnus               > Ossetic. k_oyroj “mill”

Gmc *lǣswaka           > Ossetic læsk “pasturage against payment”

Gmc *nakwina         > Ossetic læg_oyn[3] “hairless, bald, naked”

Gmc *gahwi                > Ossetic qæw “village, settlement”

Gmc. *rapaina            > Ossetic rævæjnæ “long, thick hemp-rope”

Gmc. *wīsa                 > Ossetic wis “patch with mowed grass”

Gmc. *strab-               > Ossetic sævn “width of clothes”

Gmc. *spīra-               > Ossetic fsīr “ear of corn”

Gmc *kurdra-          > Ossetic k’ord “group, much, bundle[4]”

Gmc. *spelda-             > Ossetic syvældæg “layer”

Gmc. *apul-        > Ossetic fæ-tk_oy < *æfæl- tk_oy “apple-tree”??? (highly speculative)

The same assertion holds true for language contact with Slavic. While quite some Germanic words have entered Slavic through (presumably) the Gothic language, the other way around no borrowing on such a scale has been active. Consider for example these loans:

An interesting article by Kortlandt (2001)  titled “The origin of the Goths” argues for a Central European urheimat for the ancestors of the Gothic speaking people. He convincingly argues that the Proto-Goths must have been situated more to the west than has often been assumed. The borrowing of the agentive –āreis suffix from Latin –ārius pleads for this, along with the fact that Gothic borrowed the  Latin form of most Greek ecclesiastical terminology. Supposed loans from Celtic also point to a more western origin. Kortlandt holds open the possibility that the Gothic identity was transferred from the Baltic Gutones via Wolfram’s Traditionskerne. Unfortunately, Kortlandt does not treat the aforementioned loans from Gothic into Alanic and Slavic, but they are easily explained from late fourth century contact between Goths and Slavs/ Alans, when the Goths reached the mouth of the Danube. Especially the Alans remained bound to the Goths for most of the following century, fighting alongside them at Hadrianople (378), Rome (409) and at the epic battle at Châlons (451) where the Romans and their Alanic-Gothic allies went into battle with the Huns of Attila.

This case shows how important an interdisciplinary approach is to tackling migration history. One could ask therefore why historians do not make use of linguistic arguments. One reason is probably to be found in the technical nature of the linguistic discipline. Another concerns the methodology of historians who work in a hermeneutic science and are sceptical of the positivistic approach of the comparative method of linguistics. One sometimes even hears "I simply do not believe in Indo-European" from respected and renowned historians in the field. As historical linguists we should help span the divide between old Germanic philology and medieval studies by making our arguments more accessible to the interested historian; Kortlandt’s article is a good example of this, focussing on cultural loans instead of purely linguistic argumentation.

A thing we could and should blame the historians for is ignoring the linguistic diversity of late antique and early medieval Europe, restricting themselves to Latinate sources. One need but to remember that Latin was not the only written language in the early medieval west. From the period of 400 - 900 AD we find literary monuments written in Celtic, Germanic and Slavic vernaculars. Let´s remember that only a small percentage of the population knew Latin, so the generational transfer of culture in the Early Middle Ages could only take place by use of the vernacular. When one considers the fact that early medieval vernacular sources often deviated immensely from the genres of rigid and devote Latinity, the comparative Indo-Europeanist could only sigh and shake one’s head at how many gems the historian leaves untouched. Historians should better heed the word of Jordanes himself to “follow the writings of their ancestors and cull from their broad meadows a few flowers to weave a chaplet for those who care to know these things”.

Bibliography

Jay, H. Jasanoff, “Gothic”, in: the ancient languages of Europe, Roger D. Woodward ed. (Cambridge 2008) 189-214.

Augustin Speyer, Germanische Sprachen; ein vergleichender Überblick (Göttingen 2007).

Eckhard Meineke et Judith Schwerdt, Einführung in das Althochdeutsche (Paderborn 2000)

Thomas F.X. Noble, “Introduction; Romans, barbarians and the transformation of the Roman empire”, in: From Roman provinces to medieval kingdoms, Thomas F.X. Noble ed., (New York 2006) 1-28.

Herwig Wolfram, “Gothic history as historical ethnography” in: From Roman provinces to medieval kingdoms, Thomas F.X. Noble ed., (New York 2006) 43-69.

Walter Goffart, “Doest the distant past impinge on the invasion age Germans” in: From Roman provinces to medieval kingdoms, Thomas F.X. Noble ed., (New York 2006) 1-28.

Bryan Ward-Perkins, The fall of Rome and the end of civilization (Oxford 2005).

J. Svennung, Jordanes und Scandia; kritisch-exegetische studien (Uppsala 1967).

“Gutasaga”, in: Altschwedisches Lesebuch, Adolf Noreen ed., (Upsala 1892-94) 37-39.

F. H.H. Kortlandt, “The origin of the Goths”, in: Amsterdamer Beiträge zur älteren Germanistik 55 (Amsterdam 2001) pp. 21-25.

Robert Anderson, “Foreword”, in: The transformation of the Roman World, Leslie Webster et Michelle Brown eds., (London 1997).

Han Nijdam, “Twee aardewerken schaaltjes. Collectief geheugen, (dis)continuïteit in de Friese cultuur, met het fenomeen ‘magie’ als case study”, in: Speculum Frisicum; stúdzjes oanbean oan Philippus H. Breuker, Rolf H. Bremmer Jr. e.a. eds., (Leeuwarden 2001) 59-78.

Vasily Abaev, Istoriko-jatymologičeskij osetinskogo jazyka (Moscow 1968).

[1] My former professor in Utrecht, Mayke de Jong, confessed that she’d rather speak of “Sub-Roman” traditions than of “Germanic” traditions.

[2] [2] Jordanes names is interpretated as *Iƀurnanþs by Grimm who draws for this interpretation on the spelling <iornandes> in some existing manuscripts. He argues that sixteenth century editions always read <iornandes> and may reflect additional old manuscripts that are lost. It is interesting to note that his father is called Alanoviamuth, in which the first element undoubtedly refers to the ethnonym Alani. Maybe the name should be read as *Alano-Weihamōþs “der Kampfmutige der Alanen”. Alanic descent for Jordanes is also suggested by the name of his grandfather, who was named kandag, which could go back to Old-Ossetic *kæn-dag “he who wears sack-cloth”.

[3] Regular dissimmilation of *n...n > l...n as seen in lamaz “Islamic prayer” (< Pers. namāz)

[4] Glossed by Abaev as “группа, множество, стая” in Russian

OHG <trikil> as the sole correspondance to ON þræll

OHG *drigil "house slave" is commonly considered a comparandum to ON þræll (< PNo *þraχilR) "slave "and OE ðrǣll[1].  These last two forms would reflect PGmc *þraχila-, whereas the OHG cognate would need *þreǥila-. In older etymological dictionaries also a non-existing OHG **dregil and OS **thregil are often cited. These entries into the comparative material give the impression that also a PGmc *þraǥila- existed and is reflected in OHG and OS, while in reality OHG *drigil is the only continental Germanic correspondance to the OE and ON words. Van Helten also hypothesizes *trachil from trachlagia “animal vel iumentum”  in the Malberg Glosses of the Pactus Legis Salicae but both the semantics and the obscurity of the gloss do not present convincing material for the connection. A connection of French drille[2] “soldier, vagabond” to OHG *drigil is unlikely since the word is attested really late (17^th century) and is probably loaned from ModDu drillen “to drill soldiers”.

 To get back to OHG *drigil, only Schützeichel’s althochdeutscher und altsächsischer Glossenwortschatz[3] lists <trikil>, with a closely related <trikilod> as an abstract formation to presumed *drigilōn. As mentioned before, OHG *drikil implies Pgmc *þreǥila-, which is often connected to Goth þragjan “to walk’ by use of the agentive suffix Pgmc *-ila-. Old formations in Pgmc *-ila- usually have zero-grade of the root, cf. OHG brûtpitil "paranymphus" and OHG (stein)-brukil "mason". In this regard OHG *drikil would deviate from the expected zero-grade of old formations.  The semantic development of “walker” to “slave” has a inner Germanic analogue in MLG drāvel, drevel “slave” from drāven “trotting, running”[4].

With trikil the empirical basis for *drigil, we should get back to the source where the word is to be found. This is Manuscript K, in German known as “das Keronische Glossar”, of the Abrogans (St Gallen, Stiftsbibliothek, cod. 911), with trikil as a gloss to Latin verna[5] “house slave”. Manuscript K, dated to the late 8^th century, was not written in St. Gallen, but a more precise localization other than within the South West German writing tradition is not possible[6]. The word is to be found on page 279 of the manuscript, separated from the following word by a dot and written in unconnected uncials with red colouring of the empty space in the initial letter. The Latin word that trikil glosses stands in front of it and reads <uerna> with an abreviation stroke on top of the <u>. The abbreviation sign the scribe chose is slightly wondering because following <uernat>, <uernabunt> and <uernaculus> on the same page all have a different abbreviation stroke. For lack of convincing alternatives, a reading verna is to be preferred. Concerning the graphemic form of trikil it should be noted that initial <t> instead of /d/ and medial <k> instead of /g/ are common to the manuscript. The <i> however is not a scribal variant of <e> since the scribes of Manuscript K keep the two graphemes diligently apart in their vernacular renderings of the Latin,[7] e.g. OHG hrekil for Latin trophea and not **hrikil. Therefore we cannot attribute the <i> to Romance influence and can safely reconstruct OHG *drigil .

OHG trikilod glossing liminum (=liminium) “exile” is found in two manuscripts, one being the aforementioned Manuscript K, the other one in Manuscript Ra (Karlsruhe Landesbibliothek, Aug CXI f84^ra) that belongs to the same tradition as Manuscript K, dependent on the Paris manuscript (Paris bibliotheque nationale, cod. lat. 7640). Here we find <strigilod> with an erroneous initial <s> and the expected medial <g>.

If compelled to find other (etymological) solutions for OHG *drigil besides Pgmc *þreǥila- we could consider Graff’s[8] comment on OHG trikil that Manuscript K also contains the word tripil (a variant of OHG tribil) glossing famolus (=famulus) (or maybe a latinized sculla? cf. MHG schülle?). It may therefore have been inserted into the wrong line of the glossary. However, the words stand really far apart in the glossary and other attestations of tripil have it gloss auriga “charioteer” and agitator “id.”, thus indicating that the meaning of the two words were also very different. Reading tripil instead of trikil therefore constitutes a severely flawed counterargument.

One could also consider connecting OHG trikil with OHG tragan and considering it a variation to OHG tragil, tregil “carrier”. For Indo-Europeanists this would be good news since the root PIE *d^hreǵ^h- is reflected in e-grade in Gk τρέχω. For an IE root reflecting Goth þragjan and related ON þræl we cannot find valid IE comparanda and therefore the IE nature of the root and the ablaut are not secure. The downside, however, of explaining trikil as a variant of tragil would be that formations with OHG tragil are very well attested in the gloss material and none of the attestations show **trigil.

An interesting problem is how OHG trikil / *drigil would relate to OE þrǣgan “to run” < Pgmc *þrē¹ǥejanã, since the OE word also seems to be related to Goth þragjan < *þraǥejanã. This relation is the more interesting since ON þrǣll  (see Schaffner, 2001, 167) could also be explained by *þrē¹χila-, with a corresponding analogue to the lengthened grade in OE þrǣgan. However, the Vernerized spirant of the OE verb would not correspond in the voiceless spirant of the ON noun. Pgmc *þrē¹ǥejanã would then be a denominal formation to OE þrāg /þrāh “time, course” < Pgmc *þrē¹ǥō, comparable to Pgmc *frē¹ǥō, cf. OHG fraga, OF frēge. We would expect denominal verbs from feminine o-stems to yield weak verbs of class II in Pgmc *-ōjanã, i.e. Pgmc *þrē¹ǥōjanã > OE **þrǣgian as Pgmc *frē¹ǥō → Pgmc *frē¹ǥōjanã > OHG frāgōn. Moreover, OE þrǣgan is evidently a class I weak verb considering its attested preterite OE þrǣgde 3.sg.pret. in the Elena poem (9^th century). OE þrǣgan wk.I could have developed out of an earlier wk.II verb *þrǣǥian or out of an old reduplicating verb *þrē¹ǥanã, cp. OE grǣtan < PE *grǣtijan wk.I ← *grǣtan str.red., cf. OE grātan (see Seebold, 1978, 241). To my mind it seems prudent to reconstruct an ablauting PGmc *þraχila-/*þreǥila-, for a direct derivational relationship between *þrē¹ǥejanã and ON þrǣll seems problematic.

To conclude: OHG *drigil “house slave” can be reconstructed on the basis of the Abrogans gloss trikil gl. verna in Manuscript K. It can be connected to a Pgmc agentive formation exhibiting e-/o-ablaut in Pgmc *-ila, i.e. Pgmc *þraχila-/þregila-, in which the e-grade, next to the o-grade of ON þræll is confirmed by the OHG trikil.

Bibliography

Schaffner, Stefan, Das vernersche Gesetz und der innerparadigmatische grammatische Wechsel des urgamnischen im Nominalbereich, (Innsbrucker Beiträge zur Sprachwissenschaft bd 103) (Innsbruck, 2001).

Elmar Seebold Vergleichendes etymologisches W¨orterbuch der germanischen starken Verben (The Hague, 1970).

L. Sütterlin, Geschichte der Nomina Agentis im Germanischen  (Strassburg, 1887).

Rudolf Schützeichel, Althochdeutscher und Altsächsischer Glossenwortschatz, bd 2 (Tübingen, 2004).

Elias Steinmeyer et Eduard Sievers, Die Althochdeutschen Glossen, bd 1 (Berlin 1979) 263.

Die deutsche Literatur des Mittelalters; Verfasserslexikon, eds. Kurt Ruh e.a., bd 1 (Berlin, 1978)

Etymological inquiry into the origin of ModDutch viool and vedel

 Since my native language is Dutch my interest in the etymology of specific words often but not always starts with a simple inquiry into my native idiom. This was also the case when I wondered about the etymology of Modern Dutch viool “violin”, which is of course related to ModE violin and ModFr viole. The direct etymology of the word is to be found in Middle French viole which itself goes back to Old Provencal viola / viula. The source for this OProv. viula is without a doubt Medieval Latin vitula or vidula “stringed instrument, lyre”, but here things get problematic. What is the origin of MLat vitula / vidula? The Latin word is first attested in the eleventh century (MLat vidula) and attempts to connect it with vītulāri “to exult, be joyfull” on the one hand and fidēs / fǐdǐcǔla “stringed instrument” on the other are not without difficulty.

Latin fidēs and its diminutive fǐdǐcǔla are often assumed to originate in a mediterannean substratum language that also yielded Greek σφίδες “tripe fit for cookery”. The Medieval Latin /-t-/ in vitula should then be considered a Romance confusion of voiced and voiceless intervocal obstruents, which seems to be unlikely for a twelfth century attestation. The verb vītulāri seems to be derived from the name of the godess Vitula or Vitellia, who is the godess of joy, so this verb also doesn´t give us a solid etymology for vitula[1].

The modern French word vielle, by most historians of music erroneously thrown at the same etymological heap as ModFr viole and ModE fiddle, is plausibly derived from fǐdǐcǔla, assuming that Romance /e/ < Lat /ǐ/ in this case merged with /ɛ/ that could diphtongize to /jɛ/. That ModFr vielle does not share the same origin as ModFr viole is clear from the OFr /-ʎ-/ that must come from a palatalized /-kl-/, cf. ModFr abeille “bee” < Lat. APICULA and ModFr caille “quail”< Lat. QUACULA.

When it comes to the origin of Medieval Latin vitula / vidula it seems therefore reasonable to look to that other major source of Medieval Latin idiom, namely the Germanic languages, a solution that was preferred by amongst others Warthburg and Bloch in their Dictionnaire Etymologique du Français. Here we encounter a family of words which is often assumed to have their origin in just that Medieval Latin word vitula, to which the ModDutch word vedel “medieval string instrument” and ModE fiddle belong. The earliest attestation of the ancestor of these Germanic words is to be found in Otfrid of Weissenburg’s Evangelienbuch as fidula.

Otfrid’s Evangelienbuch, V 23, 197-201

It should be noted that in this fragment líra and fidula are presumably placed together for being similar instruments in the same way harpha and rótta[2] are placed together. The word fidula  is also attested in a tenth century manuscript (Ro Pal. Lat. 1517) as fidala, glossing fidia for fidicula (confusion with tibia?) in the work of the late antique poet Prudence. However, when one looks at the vernacular glosses it should be noted that fidala and its younger reflexes do not only gloss fidicula, but also tibia “flute” and the nomen agentis fidulāri glosses tibicen “flute-player”. The identification as a string instrument is therefore not certain from the Old High German glosses. In Old English however the identification as fidicula is more certain since OE fiðelere and fiðelestre do gloss Latin fidicen “lyre-player, someone who plays a string instrument”.

Considering the Old Germanic comparanda, i.e. OE fiðele, OHG fidula, ON fiðla, it must be noted that these merit a Common Germanic reconstruction as Gmc *fiðlōⁿ . If one insists on seeking the origin of the Old Germanic words in Medieval Latin vitula / vidula, one is confronted with the very late attestion of vidula (two centuries later than the first OHG attestations!) and, more importantly, the perplexing presence of Gmc /-ð-/ for Latin /-t-/ or /-d-/. Furthermore, the word was loaned into Old/Middle Irish as fidil from Germanic, not from Latin. How to account for the formation of Gmc *fiðlōⁿ? We could be dealing with a part loan from Romance, which acquired a Germanic suffix, namely the part *fidi- from fidicula with the Germanic instrument noun suffix *-ilōⁿ. Another solution concerns onomatopaeic formations which are amongst others to be found in the Romance languages themselves. Here we find quite some words concerning music and sound that have a sound imitative origin, e.g. Provencal piular “to bemoan, to yell”, miular “to cry” and *fiular “to whistle”. The Dizionario Etimologico della lingua Italiana doesnt it hold it unthinkable that the origin of OProv. viola / viula must be sought in a formation originally meaning “lo strumento che va viu” (an instrument that makes a viu-sound). For the plausibility of a sound imitative origin for Gmc *fiðlōⁿ one should also note the etymology of ModGerm geige “violin”, which is the modern reflex of MHG gīge, derived from the verb gīgen “to make the sound gīg”.

In the same way a Germanic sound imitative root *fi- could have acquired an instrumental suffix *-ðlōⁿ or, more plausibly, *fið- an instrumental suffix *-lōⁿ (litterally denoting “an instrument that makes a fi(ð)-sound”), compare Germanic *pīpana (cf. ModDu pijpen / piepen) as a sound imitative verb which gave rise to the nomen agentis *pīp-āri. Something similar was proposed by Van Wijk who insisted on a Germanic formation going back all the way to PIE, i.e. Gmc *fiðlōⁿ < *fiðlō < PIE *piH-tleh₂, which seems highly unlikely since the PIE root *pei̯H- “to sing”is only attested in Old Church Slavonic and Tocharian, i.e. OCS поѭ, пѣти and Tocharian B (conj.3pl.) pīyam̥.

Assuming a Germanic origin for the ancestor of ModE fiddle and ModDu vedel, it becomes plausible that Medieval Latin vidula was loaned from Germanic instead of the other way around, the sound substition of */-ð-/ for /-t-/ or /-d-/ being quite common; early /-ð-/ was first romanized as /-t-/, cf. OFrnk. *friþu- > Gallo-Roman fretum, but later developed into Romance /-ð-/ before subsequently disappearing. Naturally, also Germanic /-ð-/ was romanized as /-ð-/ before disappearing, cf. OFrnk. *laðo > OFr. laon “board” and *flaðo > OFr. flaon “specific cake” (cf. ModDu vlade /vlaai).

Concerning the etymology of ModDu viool we could start from Gmc *fiðlōⁿ > OFrnk *fiðula > Gallo-Roman fitula > vidula <viðula> > OProv. viula > MidFrench viole > EModDu viole. The initial Romance /v-/ could have arisen from lenition caused by the article, cf. una  fitula > una viðula, for intervocalic /-f-/ went through Romance /-v-/ as evidenced by OFr. Estievene < STEFANU and ravene < RAFANU. This then could also be the solution for the /v-/ in ModFr vielle.

The direct Germanic development could be illustrated by Gmc *fiðlōⁿ > OFrnk. *fiðula > OLFrnk. *vidula > MidDu vedel(e). This in it’s turn would be the reason why we have the pairs fiddle / violin in English and vedel / viool in Dutch.   

Bibliography

Alkire, Ti and Carol Rosen, Romance languages; a historical introduction (Cambridge, 2010).

Beekes, Etymological dictionary of Greek, 2 vols (Leiden 2009).

Bourciez, Edouard, précis historique de phonetique française,  nouvelle collection a l’usage

des classes III (Lille 1921).

Cortelazzo, Manlio et Paolo Zolli, Dizionario Etimologico della lingua Italiana (Bologna 2004).

Gamillscheg, Ernst, Romania germanica. (Berlin:1970).

Greimas, A. J. Dictionnaire de l’ancien français; jusqu’au milieu du XIVe siècle (Larousse, 1989).

Kluge, Friedrich, Nominale Stammbildungslehre der altgermanischen Dialecte (1886).

Vaan de, M., An etymological dictionary of Latin (Leiden 2008).

Zink, Gaston, phonétique historique du français (Paris, 1986).

The OHG Isidore translation and it’s background

A friend of mine is doing her PhD on early medieval glossaries, which brings her often to Leiden University to work with facsimiles and original early medieval manuscripts, some of them containing glossaries that are highly valuable to the linguist, although linguists rarely turn to the manuscripts themselves. We had an interesting discussion on why glossaries were collected in the first place and, more importantly, the early medieval origins of vernacular translations on the continent. Somewhere in the discussion I mentioned the Old High German translation of Isidore’s De fide catholica ex veteri et novo testamento contra Iudaeos, remembering that linguistically it was highly archaic which suggested that it must have been written before the ninth century. We both wondered why in heavens name a sixth century tractate agains the Jews was translated in the vernacular in the first place, a wondering which demanded further inquiry.

As far as we both knew the Jews were a quite insignificant minority in Carolingian society (especially in the Germanic speaking parts of the empire) who mostly worked in trade enterprises and held some small but controversial privileges, although being generally discriminated against in legal matters. In this blogpost I want to discuss the main manuscript in which the text is preserved, lay out why the tractate was translated and give you a small fragment of the text, i.e. the latin original, the OHG translation and a modern English translation of the Latin and the Old High German.

                The text that we will discuss is known as De fide catholica ex veteri et novo testamento contra Iudaeos, written by bishop Isidoris Hispalensis, Isidore of Sevilla, around 614-615 as a theological motivation to a royal decree of the Visigothic king Sisebut, ordering all Jews to convert to christianity. The Old High German translation is preserved in a late eighth century manuscript, kept in the Bibliothèque National de Paris (BN lat. 2326), and a fragment of the text is to be found in another manucript, this one preserved in Vienna (ÖNB cod. 3093). Palaeograpically the text in the Paris manuscript, evidently a copy of the original, uses an orthograpical system commonly used in Murbach. The Isidor-fragment is preceded in the Paris manuscript by an OHG translation of  the Matthew gospel, another text which the editors of the manuscript called de vocatione gentium, the final part of an eigth century sermon and sermon LXXVI of Augustine. The Latin and vernacular are given in two seperate columns. Linguistically it seems preferable to place the author of the translation in Lorraine, since the language used is Old South Rhine Franconian. Metz, St. Avold and Hornbach have been suggested as possible places. Kowalski-Fahrun convincingly argued for an early Carolingian date to the translation, since linguistically the Old High German preserved in the texts is quite archaic.

                This can be illustrated by the fact that the common OHG syncope in the past participles and the preterite of the weak verbs has not yet taken place, certain unique OHG forms are used that only have a corresponding form in Gothic (OHG lyuzilla cf. Goth luttila, OHG chillaubin cf. Goth. gilaubi, OHG mittingart cf. Goth midjungarda, OHG geizzsi cf. Goth. gaitin) and certain archaic morphological traits, such as the preservation of the u-declination, the m-ending in the first singular of the present tense, as seen in bim and sculim, and the dative plural ending in –m.

With an early dating of the text corroborated by linguistic arguments, placing the translation in the beginning of the reign of Charlemagne (770’s AD) seems reasonable, although most vernacular writings are connected to late Carolingian scholarship in the closing years of the eighth and beginning of the ninth century.  Even with a late dating of the translation, the OHG Isidore, together with the Monsee-Wiener-Fragmente, would be the oldest witness to Carolingian theological writings in the vernacular.

It is believed that the reasons for the translation of the Latin text in the vernacular were purely practical. For Isidor had collected in his tractate exhaustively all the passages of the Old Testament that had any connection to Christ and he gave an excerpt of all material from the Old Testament, with the exception of the psalms, that was necessary for the support and understanding of the christian faith. It was, as it were, a survey of messianistic prophesies along with the necessary explanations historically ordered. It would have functioned as a practical teaching manual while the original intentions of the sixth century author were pushed into the background. This is corroborated by the fact that Latin iudaeos is translated by unchilaubun, i.e. “infidels” or “unbelievers”, showing that the identification of the antagonists as Jews against whom the tractate was originally meant, was secondary.  The sixth century work contra iudaeos presented to the early Carolingian clergy the quintessence of Old Testamentic writings relevant for a proper understanding of the true faith and it is probably there where we have to look for the reasons behind the OHG translation. Remarkable is the fact that the Latin is translated in what looks like mostly genuine OHG syntax. Also, in a lot of places the OHG is a very loose translation of the Latin, aiming to convey the meaning of the text, not so much the form. The Tatian translation with its word-for-word glossing style is in many ways a step back from the high-level translating skills shown in the OHG Isidore.

                To give you an idea of what the text looks like here a fragment of the OHG Isidore translation, the Latin and the OHG, to which I added an English translation of the Latin and an English translation of the OHG. Personally, I find the late antique and early medieval logic in theological matters that this text conveys highly amusing and risible. 

Latin version

4. Si Christus deus non est, dicant Iudaei nobis, quem sit affatus deus in Genesi cum diceret: ‘Faciamus hominem ad imaginem et similitudinem nostram.’ Sic enim subiungitur : ‘Et creavit deus hominem ad imaginem et similitudinem dei creavit illum’. Quaerant ergo quis deus creavit , aut ad cuius dei imaginem condidit hominem quem creavit.

5. Quod si respondeant : ‘ad angelum,’, num angelus aequalem cum deo habet imaginem, dum multum distet imago creaturae ab eo qui creavit. Aut numquid angelus cum deo potuit facere hominem ? Quod ita existimare magnae dementiae est. Cui ergo dicitur, aut ad cuius imaginem conditus homo creditur, nisi ad eius, cuius una imago cum deo est et unicum nomen divinitatis est.

English translation of the Latin

4. If Christ isn’t god, as the Jews may say to us, to whom would God have spoken in Genesis when he says: ‘Let us make man in our image and our likeness.’ Because it is joined with the following : ‘And god created man,  in the image and likeness of god he created him’. Thus they may ask which god created or in the image of which god he composed the man that he created.

5. But if they may answer: ‘in the image of the angel,’ does the angel not have an image similar to god, while the image of that being is very different to him who created.  Or surely the angel couldnt make man together with god? To believe this is a sign of great folly. To whom is it said thusly or in the image of whom is man believed to be composed, unless in the image of him, whose one image is with god and is the sole name of divinity?

Eighth century Old Rhine Franconian rendering of the Latin

4. Ibu Christ got nist, sagheen nu dhea unchilaubun uns, zi huuemu got uuari sprehhendi in Genesi, dhar ir quhad : ‘duoemes mannan uns anachiliihhan endi in unseru chiliihnissu’. So dhar auh after ist chiquhedan : ‘endi got chiscuof mannan anachiliihhan endi chiliihhan gote chifrumida dhen’. Suohhen dhea nu auur, huuelih got chiscuofi, odho in huuelihhes gotnissu anachiliihhan mannan chifrumidi, dhen ir chiscuof.

5. Ibu sie antuurdant endi quhedant : ‘in angilo’: Inu ni angil nist anebanchiliih gote? Dhanne so dhrate mihhil undarscheit ist undar dhera chiscafti chiliihnissu endi dhes izs al chiscuof. Odho mahti angil so sama so got mannan chifrumman ? Dhazs so zi chilaubanne mihhil uuootnissa ist. Huuema ist dhiz nu zi quhedanne odho zi huues chiliihnissu uuardh man chiscaffan, nibu zi dhes dher anaebanliih ist gote endi chinamno ist mit godu ?

English translation of the Old High German

                4. if Christ isn’t god, the unbelievers may say to us, to whom is god speaking in Genesis, where he says: “Let us make man similar to us and in our likeness.’ And where also is said after that: ‘and god created man, similar and liking to god he created him. However they may ask, which god would have created or in the likeness of which god he would have made man, when he created him.

                5. If they may answer and may say: “in the image of the angel’: is the angel not similar to god? Because there is a very big difference in the likeness of that creature and of him who created all. Or would the angel and god have created man together? To believe this is great folly. To whom should this be said or in the likeness of whom is man created, if not of him who is similar to god and who is the namesake of God.

Bibliography

Wilhelm Braune, Althochdeutsche Grammatik (12th edition 1967: Tübingen 1868).

Braune, Wilhelm, Althochdeutsches Lesebuch (17th edition 1994 Tübbingen; 1875).

Eckhardt Meinecke et Judith Schwerdt, Einführung in das Althochdeutsche (Paderborn 2001).

Herta Kowalski-Fahrun, “Alkuin und der Ahd. Isidor”, Beiträge zur Geschichte der deutschen Sprache und Literatur 47 (1923) 312–324.