I.1 The corpus of the ChLA I–II and CXVIII

The volumes of ChLA I–II and CXVIII contain Vulgar Latin parchment charters from the 7th century to 801 related to St. Gall or to its vicinity.⁸ At the beginning of the first volume, there are late antique papyri fragments that were excluded from the analysis due to their dates. Parchment fragments (mostly containing a few words) were also left out, even if they contained determiners/demonstratives since the referents of the demonstratives could not be identified (ChLA I 1a,b–39, ChLA II 178). At the end of the second volume, letters (ChLA II 171–177) and vows can be found differing from the general type of the documents, from the donations and precaria. This volume also includes charters written in the chancellery of Charlemagne (ChLA II 157–158: these documents also have Vulgar Latin tendencies). In order to create a homogeneous corpus, it was necessary to exclude these peculiar documents on the basis of different style or different place of origin. On the other hand, the ChLA CXVIII is a supplement edition, therefore, there are several charters in it not just from St. Gall. For the aforementioned reason, only the documents of St. Gall were included in the examination (ChLA CXVIII 22, 24, 27, 39). Copies, even from the 9th century, were excluded on account of the stated time-span (ChLA I 42, 47, 53; ChLA II 129, 137, 160; ChLA CXVIII 1–21, 23, 25–26, 28–38, 40–46). That way, a homogen corpus of 133 charters was created, consisting of mainly donations and precaria.⁹

I.2 The common structure of the charters of the ChLA

In this paragraph, first, I intend to define what a formula is, then to describe the structure of the legal documents of the ChLA.

Due to the rigid wording of the charters, formulas can manifest themselves in short syntagmas (ad ipsum monasterium, in ea vero ratione, id est) forming sentences containing special legal terms (dono, transfundo, commutatio/concambium, proservire, precaria, visus sum habere). Apart from technical terms, there are several redundant elements to specify and identify the important entities mentioned previously in the charter and to avoid misunderstanding: frequently used adnominal demonstratives forming direct/indirect anaphoric chains, repetitive sentence compositions resulting in textual framing. Therefore, special syntagmas and the rigid wording of sentences establish the distinct structure of the texts and create four main parts: clausula donationis ending with census, clausula poenalis, clausula firmationis, finally, the closing unit containing the signatures, the date and place of the composition, the scribe's name.¹⁰ Thus, under the name formula, actually different units of the text can be understood: syntagmas, clauses and the sentence itself as a result of the combination of these clauses and syntagmas.

As it can be seen, the act of donation is expressed three times, the beneficiary is mentioned four times. Precision is achieved through determiners (in vila … in istas villas) or repetition (ad ecclesia sancti Gallonis). The listed possession is expressed through an explicative hoc est.

The referential chain is maintained via demonstratives as the text in bold shows it. To avoid misunderstanding and misinterpretation, the heirs of both genders are stated separately so that they would fall under the conditions of the contract. Double determination can also be seen that is a noun phrase containing a noun and a demonstrative modified by a relative clause.¹⁴

In respect of presens donatio istic, the formula has variants in which iste or hic alone or combined with praesens is able to express closeness in space and time. On the other hand, stipulatione subnixa is a rigid phrase that also lost its original form and meaning, nevertheless it is commonly used by scribes due to the strict tradition of charter composition.¹⁵

From this description, it can be seen that the wording of charters differs from the classical norm not just in the legal terms and repetition but in the special, frequent use of demonstratives marking the referential chain in order to avoid misunderstanding. Due to the rigidity of the formulas, the question arises how it is possible to conduct the research. Precisely, this fossilised language and its demonstrative-specific syntagmas help to assess the standard form (past tradition) of each formula and when comparing them to the deviations (innovations), the outliers manifested in determiner multiplication, change or additional metalinguistic participles, the differences give the result. This means that the standard form of a formula can be compared 1) to its variations or 2) to formulas containing the same determiner.

As it can be seen, the pia causa is mostly composed with the help of short syntagmas and whole clauses. The Biblical passages are introduced in the same way and only a few of them are cited in the charters. Variety is achieved by combining syntagmas with clauses of Biblical citation and by using more/less syntagmas with different sequence. The last one is a peculiar case, because it inserts the reason behind testament writing into the pia causa. This insertion is actually a narratio, however, these parts occur scarcely.¹⁶

To show what I regarded as a presumably erroneous formula, the formula labelled as sanctus can fulfil this purpose. This is the ubi-clause that describes the beneficiary at the beginning of the clausula donationis. Three major forms can be detected:

1. 1)ubi eius/ipsius sacrus requiescit corpus (in Christo)¹⁷
2. 2)ubi ipse in corpore requiescit¹⁸
3. 3)ubi ipse sanctus in corpore requiescit (ChLA I 90. 1) and ubi ipsi domnus in corpore requiiscit (ChLA I 84. 1)

In all examples, eius/ipsius, ipse (sanctus/domnus) refers to Saint Gall. The first type (example 1; Σ 19) contains eius (Σ 17; 5 of the total stands after the word corpus) and rarely ipsius (Σ 2; 1 standing before, 1 standing after the noun).¹⁹ Regarding the second variant (example 2), all the ten occurrences have the same structure. The two anomalies are shown in the last example. All of them can be valid forms of formulas, however, the rarity of ChLA I 90. 1 and ChLA I 84. 1 can also suggest either contamination of the first form (sacrus) with the second one (ipse in corpore) or other, unknown origin. Whatever is the cause, the demonstratives have different linguistic features in each type.²⁰

I.3 Methods applied to categorise and assess the data

The following determiners are examined in this paper: is, hic, iste, ille, ipse, idem and metalinguistic participles (supradictus etc.) on account of their functional overlap. To gain information from the charters, there are two main ways to categorise the data.

I.4 Reference of the definite article and the demonstratives

To describe the process from demonstratives into articles, many hypotheses have covered different aspects of it. To name a few: grammaticalization, semantic bleaching,²¹ intersubjectification²² and defocalisation.²³ However, these theories are hard to apply in languages without native speakers like (Vulgar) Latin, since semantics of an adnominal demonstrative or grammaticality of a sentence/syntagma cannot be determined with personal feelings about it, only indirect conclusions can be made. This is even harder, since definite articles originated from demonstratives and the same morphology covers a set of different semantics.²⁴ Both of them are part of the same category (determiners), can be used in deixis, ana- and cataphoras with semantic difference and can fulfil the role of the third person pronoun.²⁵

In the evaluation of articuloïde cases, the following criteria had been set: first mention use,²⁶ obligatory usage,²⁷ the entity marked by the definite article is part of the society's or the humanity's broader knowledge. Therefore, the analysis of anaphoras can be of help in the determination of (in)directness. The more indirect it is, the more abstract is the connection between referent and anaphor. I applied the four categories of Skrzypek et al.²⁸ The first one (1) is the deictic role, the immediate situational use when the referent is present (in the actual space and/or in the text). In the previously described charter, its equivalent is presens donatio istic. (2) Direct anaphora: the word after first mention is retrieved through additional demonstratives: in vila … quid in istas villas visus sum habere. (3) As regards indirect/associative anaphora, the antecedent is not spelled out, the context, entities of the discourse serve as an anchor. Therefore, the trigger is not necessarily a nominal phrase. For instance, a meronymic (whole-part) type: omnes facultates suas per cartam traditionis firmavit … (list of monks et priest) … diebus vitae suae annis singulis inde censum solvat de ipsa terra, in loco, qui dicitur (ChLA I 55. 1–5). It is worth noting that modern languages which have articles, prefer and use articles in this case.²⁹ (4) The larger situational use depends on the knowledge of a society or humanity and expresses uniqueness (the Queen, the sun).

Before the analysis, for a better understanding, the methodological frames of each embedded analysis will be outlined. After chapter II, in which a brief comparison is given between the data of the ChLA and the Fontes Iuris Antiqui [FIRA] II, chapter III deals with the reduction of is. It is based on both formulaic (high number of id est before lists) and linguistic notions (the disappearing neuter). The formulaic method is introduced within this frame. In chapter IV, the semantics of ipse is examined in all formulas that are not as complex as res or have so little ipse in them as charta. Again, the linguistic categories (anaphora, intensity) set the frame of the investigation for the formulas and rare variants. However, the frame changes in chapter V and VI. In chapter V, the umbrella category, res is analysed separately due to its complexity. The textual functions of several determiners (hic, iste, ipse, ille) are examined within the formulas, thus, the linguistic method is embedded into the formulaic one while further formulaic elaboration shed light to the nuanced ‘behaviour’ of the demonstratives. Chapter VI follows the structure of chapter V, although, with less complexity. Chapter VII is the syntactic evaluation of demonstratives in formulas, rare variants and also narrationes. Chapter VIII examines article-like cases. Again, linguistic methods are simultaneously applied with the formulaic one.

IV.1 Ipse as the general anaphor

In this section, the main function of ipse, its adnominal anaphoricity will be examined. The determiner ipse is well-attested in the text in different formulas,⁴³ most strikingly in the formulas labelled as monasterium, for they can directly retrieve the beneficiary of the donations, the monastery of St. Gall wherever it stands in the text:

As it can be seen in Table 6, the formulas labelled as monasterium are rigid in their demonstrative usage as they are highly ipse-specific. In these formulas, demonstratives act as adnominal possessives (Σ 27: Σ 24 eius; Σ 2 suus; Σ 1 ipsius) and as anaphoric determiners (Σ 230, only 3 of them are pronominal containing is, hic, ille, therefore their presence is almost non-existent). To retrieve the referent, adnominal metalinguistic participles also play an important role. Based on the data, it can be concluded that ipse and supradictus can retrieve the referent regardless of distance.

However, multiplication occurs in these formulas, as well. It can be found 19 times in my corpus, its general structure is either D supradictus N (Σ 15; 14 of them contain ipse, one, on the other hand, has hic) or ipse N supradictus (Σ 4).⁴⁴ How can this be interpreted? If it is correct that supradictus and ipse have the same anaphoric function, three explanations can be given: 1) ipse's anaphoric strength has undergone weakening and it is strengthened with the help of an additional supradictus; 2) ipse functions here as an identity (or intensity?) determiner, so its anaphoricity is secondary, therefore additional supradictus does not result in functional accumulation; 3) ipse is slowly developing hic-like deictic features,⁴⁵ so supradictus does not endanger its anaphoric strength (for further information, see V.1 and VI.2).

Considering the strengthening anaphoric force of ipse in the cases of id ipsum, and the fact that the other strong anaphoric determiner, ille do not occur in combination with ipse (*ipse+ille+N) to stabilise its function, the second and third option seem more feasible.

IV.2 The problem of intensity and exclusion

Since exclusion is a subcategory of intensity with extended semantics (‘no other than’), I examine them together. Before the analysis, it should be understood how intensification works. Intensification is regarded as the opposition between focus (the object marked by ipse) and periphery where the contrastive force is established by ipse.

IV.3 Overlap of anaphora and intensity

After the donator's (Adalmann's) signature, the witnesses' follow. It is the only variant of the formula where adnominal ipse stands before the name of the donator, therefore the likelihood that ipse was triggered by qui, is quite low. Inclusive intensity can also be questioned, since it is not remarkable or unexpected that the donator signs the document. The contrast between Adalmann and the periphery (witnesses and scribe) can also be ruled out: there is no same/similar name. The only reason to label it as intensifying, if he – unlike most of the donators – can actually write down his name. Nevertheless, there is another explanation, the anaphoric retrieval of the referent with(out) the function of identity. This notion is strengthened since his name occurs at the end of the clausula donationis like that: quicquid in predicta villa … ego iam dictus Adalman possessor esse videor.⁵⁰ Therefore, one has to be cautious especially around second mention names.

After the analysis, it is obvious that ipse regardless of its place (before nouns or after pronouns) can be an intensifier since ipse itself and its emphatic value trigger this function even if the periphery is not explicitly expressed in the text, only hypothetically established. Therefore, it is easier to conceive logically that ipse acts as an intensifier than as an anaphoric/identity determiner. Nevertheless, the examination of rare variants (ipse Dominus, Ego ipse … Bernegarius) suggests that intensifying ipse preserved its Classical Latin features while establishing the adnominal direct anaphoric one represented predominantly in the formulas of monasterium.

V.1 Textual functions of ipse and ille

The first excerpt illustrates the general structure of the clausula donationis resulting in census after the word [recipiam]: usually, it has a preliminary part (hoc est, quod donamus), contains framing rooted from the repetition of (in)variable formula clauses (quidquid ibidem visi sumus habere; quod ibidem visi sumus habere; que in hac die presenti [visi sumus habere]); and a generic list of the items of the land that can be expanded by adding more personal entities, in this case, the names of serfs. The list usually ends in resumptive anaphora (hec omnia) and the whole clause including the bestowed items are mainly retrieved in the charter with the help of ipsas res. It is easily observed that ipsas res_2-3 can be interpreted as the direct anaphora of ipsas res₁, however ipsas res₁ is either the direct anaphora of haec omnia or the resumptive anaphora of the bestowed entities, the constituents of the generic list.

1. (1)Propterea donamus atque transfundimus pratum unum ad carradas XII in pago, qui dicitur Purihdinga in villa Dirboheim et in alia villa, qui dicitur Speichingas, de terraturio nostro solum unum et iurnales XII et mancipiis III· his nominibus: Reginfrid, Uro, Uuinburcad ad monasterium sancti Gallonis … ipsas res possedendo censum solvere liceat (ChLA II 125. 3–7)
2. (2)Et hoc est, quod dono in Linzgauia, situm in Gaerrinberg, in loco nuncupante Hounsteti, id est casa cum curte clausa vel quicquid deinceps elaborare potuero in ipso loco trado ad monasterium sancti Galloni confessoris seu monachis ibidem deo deservientibus, habeant, teneant atque possedeant … in ea vero ratione, ut tempus vite meae ipsas res ad me recipiam. (ChLA II 119. 4–8)

The resumptive anaphoric role of ipsas res is strengthened in the examples above where there is no haec omnia to summarise the components of the generic list.⁵³ Therefore ipse+N competes with hic whose resumptive anaphoric function is mainly preserved in the short syntagma haec omnia (Σ 27).⁵⁴

Although ille is excluded from the resumptive anaphoric encapsulation, it can maintain the referential chain regardless of distance as well as ipse. However, anaphoric ille is surpassed by ipse (Σ 36: Σ 163).⁵⁵

V.2 Functional overlap between hic and iste

In the previous part (V.1.) it was mentioned that hic as a resumptive anaphor occurs mainly in the fossilised form of haec omnia (Σ 27). However, it should be also noted that hic never appears alone when functioning as an adnominal anaphor (Σ 7). It is either followed by metalinguistic participles (Σ 5; ChLA I 41. 13; I 107. 7; II 132. 8; II 138. 4; II 159. 9) or with numbers (Σ 1; ChLA I 72. 2–3)⁵⁸ and in one instance, even with idem (Σ 1; ChLA II 146. 3: haec idem loca). This feature can be interpreted two ways: 1) hic as an anaphor has undergone weakening, therefore its anaphoric role should be reenforced or 2) hic acts as a deictic, therefore the anaphoric functions are expressed with the help of supradictus etc. In chapter VI, the deictic weakening of hic can be observed, therefore the second option can be excluded. Only the anaphoric weakening of hic has been left which is somewhat attested in this chapter, since iste and ipse(+N) start to fulfil the resumptive anaphoric roles. This conclusion also entails that the additional metalinguistic participles are the signs of the evolving, but not fully established textual functions of iste (and in some cases, of ipse), as well.

To sum it up, iste, ille, ipse and hic co-occur in phrases concerning properties. However, regarding ille and ipse, distance from the referent is irrelevant, as for hic and iste, closeness to the referent is observed. Moreover, ipse is two-fold, it shows similarities with ille regarding direct anaphora, while in the phrase ipsas res, it starts to infiltrate the resumptive anaphoric functions of hic.

VI.1 The case of hic and iste

The data in Table 8 show that the majority of anaphor and all pronominal cases are attributed to is in the formula: eam infrangere voluerit. Opposed to the weakening of id, there are no pronoun multiplications in this formula, even though the word-final -m was not pronounced. The other two anaphoric cases (both in ChLA II 128. 1–5) refer to donations of the past whose contracts are changed from the donations to precaria. However, in this case, ipse can also maintain the anaphoric chain as usual, since these occurrences are second mention.⁵⁹ Spatiotemporal distance is expressed via ille referring to past and future donations, too.⁶⁰

The formulas of contra hanc donationem and qui hanc chartam fieri rogavit are evaluated together on the basis of their similarities expressing hic et nunc deixis, both situational and textual. Since both of them contain determiner duplications, the structure of the phrases are shown in Table 9:

The majority of duplications is represented by hic+iste+N regardless of their order (29/Σ 76), then hic+praesens+N follows (27/Σ 76). But the presence of iste+praesens+N is also prominent (20/Σ 76). Furthermore, there are two triplexes: praesens hic N iste (ChLA II 112. 12) and hic praesens N iste (ChLA I 76. 11).

The occurrence of temporal adjectives can be detected in formulas indicating the day of the contract: ab (hac) die (praesenti) (Σ 42). The dominant variant is praesens+N / N+praesens (Σ 25; 60%). The form hic+N is one fifth of the whole (Σ 8, 19%). The remaining 21% is the combination hic+N+praesens (Σ 9). In this formula, it can be observed that hic alone cannot act as the reference of temporally close events, since praesens or combination with it surpasses hic and the Romance outcome is the disappearance of it.⁶¹ Considering that iste starts to take over the roles of hic referring to spatially close entities in the formulas labelled as res, the deictic weakening of hic can be concluded.

Therefore, it is not a surprise that hic is giving way to iste in the formulas labelled as charta, too. After subtracting the number of cases containing determiner duplication from the total while also considering triplexes accordingly, the distribution is the following: hic+N (157/282; 55,7%); iste+N (42/282; 14,9%); praesens+N (5/282; 1,8%); duplexes and triplexes (78/282; 27,6%). The distribution is dynamic, the majority of variants is represented by hic, however, it is reinforced by praesens or iste in less than half of the cases. Between iste and praesens, there is only a little difference in favour of iste. As praesens N iste suggests, the pointing value of hic has weakened and is transferred to iste while praesens expresses temporal proximity. Apart from hic N iste, praesens N iste also shows that iste is not fully developed to completely replace hic. The weakening of hic has morphological roots, silent h- at the beginning and disappearing -c from the end of the word. On top of that, ae-e-i fusions contribute to being confused with the weak is.

VI.2 Deictic ipse

In the ChLA, 6 occurrences of ipse can be interpreted as deictic. The first two cases stem from the contamination of two formulas.⁶²

1. (1)Ut si me placuerit ipsam tradicione redemere, tunc liceat mihi cum X solidis redemere (ChLA II 136. 8–9)
2. (2)Et si hoc mihi conplacuit, quod ipsam tradicionem redemere voluero (ChLA II 140. 8)

The original word, traditio standing with hic/iste, expresses situational and textual hic et nunc deixis. In these cases (1–2), ipse substitutes iste, however the situation (the act of real-word donation and the document itself) is unchanged, therefore ipse is deictic.

The following two instances are rare formulas in which the donator guarantees that (s)he will not sue for the bestowed property:

1. (3)ego nec heredis mei nec nulla oposita persona de ipsa commutacione nihil dicere nec repedire dibiamus, sed commutacio istic omni tempore firma et stabilis permaneat stibulacione subnexa (ChLA II 117. 6)
2. (4)ego nec nullus de generatione mea nec nulla opposita persona de ipsa tradicione nihil dicere nec repetire non debeamus (ChLA II 127. 7–8)

As it can be seen, these phrases start with the ego ipse phrase of the clausula poenalis and result in the clausula firmationis. In both of them, the type of transaction usually stands with hic/iste⁶³ suggesting that the situation is deictic where commutatio and traditio have both situational (the legal act and the ritual around it) and textual connotations.

1. (5)solvat dublum tantum, quantum in ipsa donatione insertum est (ChLA I 90. 11)

Example (5) is a rare variant to set the penalty, which is usually expressed by adnominal iste.⁶⁴

In all examples (1–5), the deictic value of ipse can be deducted from its environment: ipse emerges in formulas and in contexts, where hic or iste is preferred. It has to be also noted that all the instances above are from different scribes who denote the manifestation of the legal procedure (the documents and the types of agreements) with the help of hic and iste in the same charters. However, it also has to be mentioned that the most substantial evidence is absent: determiner multiplication preferably among the formulas of charta (*hic+ipse+N).

The last example is actually an overlap between intensifying and deictic ipse:

1. (6)In ipsa casa (= ecclesia sancti Galli) fuit ipsa carta levata. (ChLA I 110. 10)

This sentence alludes to the symbolic legal act when the donator picks up the charta placed on the earth and hands it over to the scribe to write down the text on it. Ipse can express exclusion,⁶⁵ since other objects establishing the periphery (ink-well, pen) can also lie on the earth to give them to the scribe, however mentioning only the charter is enough for the execution.⁶⁶ On the other hand, ipse can have deictic value, since this phrase describes the transaction itself preceding the composition of the text. Therefore, situational deictic and intensifying ipse overlaps.

What could be seen in the analysis, is the slow decrease of hic due to its phonetic, semantic and functional weakening in favour of iste and praesens. Although the distribution is diverse, the strict tradition of fossilised formulas cannot maintain the role of hic. Apart from that, there are cases of ipse (Σ 6) expressing textual and situational deixis, however the evidence is only secondary due to the lack of determiner multiplications in deictic formulas. The reason why it is mentioned, stems from the observation of chapter III and V.1. where ipse(+N) occurred in the role of resumptive anaphoric hic suggesting that the two demonstratives share similar properties.⁶⁷ The forming deictic shade of ipse is also important in understanding the additional metalinguistic participles (suprascriptus etc.) occurring in direct anaphoric instances (chapter IV.1.), since this feature shows close similarities with hic and iste both occurring with additional metalinguistic participles when (resumptive) anaphora is expressed. The similarities yet point to different evaluations: while hic is weakening, the deictic value of iste and somewhat of ipse is evolving.

VII.1 Adnominal demonstratives without possessive functions

Table 10 represents the main formulas appearing in the present study. A (= ante) denotes the syntactic position before nouns/pronouns, P (= post) denotes the syntactic position after nouns/pronouns. As it can be seen, ipse apart from one instance stands before nouns and after pronouns (just like idem and is, although with considerably less occurrences). Hic also prefers the prenominal position, however the place of ille and iste is less restricted, the latter is mainly postnominal.

But strict positions can be maintained due to the rigid wording of the formulas. Therefore, the analysis had to be expanded on rare variants of formulas and non-formulaic, free parts of the charters. The results are listed here:

Ipse:

1. (1)ipsi domnus in corpore requiiscit (ChLA I 84. 1)
2. (2)ipse sanctus in corpore requiescit (ChLA I 90. 1)
3. (3)ipse Uuigant mortuus fuerit (ChLA II 144. 10)
4. (4)ipse Dominus (ChLA CXVIII 24. 1. 1)
5. (5)ipse nepos meus (ChLA CXVIII 27. 11)
6. (6)misso ipsius Geralti comitis (ChLA CXVIII 39. 4)
7. (7)de ipsu agru (ChLA II 155. 3)
8. (8)qucumque habuit apud se ipso ad ipsos sanctos tradidit (ChLA I 51. 5)
9. (9)me ipsum Teotfridum (ChLA I 61. 4)
10. (10)me ipso (ChLA I 88. 4)
11. (11)quicquid ibidem tu ipsic et frater tuus per cartam tradicione ad monasterio sancti Galluni fecistis (ChLA II 165. 3)

It can be observed that ipse is prenominal (1–7), but always stands after personal pronouns (8–11). What is more, these findings also give an explanation to the structure of the id-ipsum. So, the general structure for nouns is ipse+N, for pronouns is D+ipse.

The other demonstratives appearing in rare variants and narrationibus:

1. (12)illi pagesis cumiatum habeant (ChLA I 55. 7)
2. (13)illi hominis vestri (ChLA I 95. 15)
3. (14)Et illi bresbitero donat illo Herigero (ChLA II 131. 13)
4. (15)hanc licentiam redemendi non habeat (ChLA II 144. 9)
5. (16)isti homines sunt secularii adscripti (ChLA I 92. 8)
6. (17)iste Cunzo (ChLA I 101. 7)
7. (18)seculum istum derelinquam (ChLA II 143. 10)
8. (19)pro beneficias eiusdem feminae (ChLA I 100. 3)

Examples (12–14) represent instances of ille, all of them prenominal, although, the number of occurrences is low just like in the cases of hic (15) and idem (19). Both of them correspond to their syntactic behaviour of the formulas. In the freer parts, the position of iste (16–18) is free just like in the formulas. It can be concluded that the syntax of freer parts converges to their formulas. When determiners multiplicate in a syntagma, the determiner reenforcing the original function stands after the weakened one (see Table 9 for iste and chapter III for ipse).

VII.2 Demonstratives as possessives

The formula labelled as sanctus was analysed in depth in chapter I.3. The other occurrences (Table 11) mainly belong to 1) the property list when serfs are given together with their land and family and 2) the not compulsory description of the kinship of the donator. In the narration, only one case was found: a sociis ipsius (ChLA II 161. 5). It attests for the postposition of ipsius found both in the formula sanctus and in other parts of the charters. Illius, although less frequent, prefers standing before the noun. Eius prefers prenominal positions, however, it is also attested after nouns. Its proportion 3:1 is almost the same in sanctus as in the freer parts. On the other hand, suus are somewhat closer to ipsius due to its frequent postnominal occurrence. What can be stated: eius even in the same formula of sanctus does not influence the position of ipsius. In fact, the position of ipsius is similar to suus.

To recapitulate the main features of the demonstratives' syntactic position, non-formulaic parts follow the positions of their respective formulas. However, the small amount of illius and of is makes the comparison insufficient. On the other hand, ipsius deviates from ipse preferring the postposition, while the position of ipse is restricted standing after pronouns and before nouns. Iste alternates freely, however, when appearing together with adnominal hic, it is placed after hic or after the noun following hic. Interestingly, in determiner multiplication, the compensating determiner always occurs after the weakened one.⁶⁹

VIII.1 Agreement changing charters

In all cases, the donated possession is regarded as illas res (those things) modified by a relative clause describing the past act of the previous donation. In what follows, I intend to argue that the illas res, quas… tradidit formula fulfils the linguistic criteria of demonstratives of recognitional use while establishing the (cultural) context of the agreement changing charters.

Regarding recognitional use, first time mention adnominal demonstratives (in case of Latin, it is ille) can be used to retrieve shared knowledge between hearer and speaker. Additional information conveyed by relative clauses can also be present in the discourse, however, it is not compulsory.⁷¹ Three criteria are to be fulfilled in these cases: the information should be new to the discourse (first time mention), old to the hearer and private.⁷² It is irrelevant, whether the actual referent can be seen or it (re)appears only in the mental representation of both parties.⁷³

How can contract changing charters meet the linguistic requirements? It is known that charters have more copies.⁷⁴ At least two copies should be made: one for the beneficiary, the other for the rogator/donator. That way disputes over ownership or shares can be solved with the help of these documents or by the summoning of the witnesses whose signatures the charter contains.⁷⁵ Since it is unlikely that the monastery would give the right of usufructuary to a nescio quis, charters are the legal basis of any claims intended to modify the original agreement. Donations mostly contain additional clauses for repurchase or for the future usufructuary of the precaria. These rights can be inherited, when stated in the charter.⁷⁶ That way, the donator's (grand)children can ask for precaria or repurchase, too, even if the whole inheritance was granted.⁷⁷ To conclude: the charters were preserved in the monastery, as well, therefore both rogator and the monastery share the same knowledge of the past transaction and what it entails. The fact that the contents of the charters are known are proven with charters containing no location at all, only illas res.⁷⁸ It is obvious that the examples above show first mention adnominal ille modified by relative clauses. After establishing the cultural context with the help of actual charters, it can be also conceived that the past donation is known for both parties. Therefore, the conditions, new to the text, old to the hearer (to the rogator) are fulfilled. The charters of the ChLA I–II and CXVIII are private legal documents, as it is stated in I.1. The content of the agreements is known to the parties (and presumably to their acquaintances), but not publicly disclosed to Raetia. Therefore, the linguistic conditions are satisfied.⁷⁹

Lastly, it is hard to decide whether illas res, quas is a Vulgar Latin innovation or a Classical Latin inheritance, because previous charter examinations of demonstratives do not distinguish between recognitional use of adnominal ille modified by a relative clause and cataphoric ille, therefore recognitional use of ille is hidden among the cataphoric cases. With that in mind, these cases alluding to previous legal agreements are certainly not articuloїde.

VIII.2 Articuloïde ipse?

The church – as usual – is defined by a relative clause describing the location. Ipse appears before the Salic land first, however a few lines later, it does not occur in the same phrase, although its reappearance would be stipulated for the maintenance of the anaphoric chain. As I mentioned in chapter IV, adnominal ipse can be an exclusive intensifier, especially if it appears at first mention. It cannot be stated whether ipse is an article or a demonstrative, since its exclusive intensifying semantic cannot be ruled out. However, the lack of ipse at second mention (which can be also understood as a scribal error), before the same noun contradicts the assumption that first mention ipse is an article here, since the referent is not precisely retained. Therefore, it is very likely that ipse acts as a demonstrative here.

VIII.3 Exceptum phrases

De illa ecclesia portionem suggests the same possessive interpretation, however ille stands before ecclesia referring directly to the same monastery that is the beneficiary of the contract. When contrasting the two phrases, ille can be interpreted three ways: 1) scribal error that placed ille before res and not before ecclesia; 2) ille expressing opposition between the donated property of the charter marked by ipsas res and the previously acquired possession of the monastery marked by ille;⁸¹ 3) articuloїde. Since the third possibility can only be proven with more contrastive evidence and first and second explanations cannot be ruled out completely, it is unlikely that ille is articuloїde here.⁸²