Context

The Church of the Holy Sepulchre is a large complex of structures situated in the Christian Quarter of Jerusalem's Old City. It was initially constructed during the reign of Constantine I in the 4th century atop the remnants of demolished Hadrianic public structures. Over the centuries, it has faced numerous cycles of damage, rebuilding, and even complete redesign. Most of the surviving structures date to the Crusader period. At the eastern edge of the church complex lies the Chapel of St. Helena, currently under the custody of the Armenian Patriarchate. In 1970, the Patriarchate initiated excavations in the chapel to investigate whether there was an empty space or bedrock behind the apses of St. Helena. The excavation revealed a space that had been blocked since antiquity, now recognized as the Chapel of St. Vartan.

In 1971, the Armenian Patriarchate invited Archie Walls and Sven Helms from the British School of Archaeology in Jerusalem to oversee the excavation. Inside the chapel, they discovered a smooth-faced ashlar that displayed a drawing of a merchant vessel along with an inscription (Fig. 1).

For the inscription in its archaeological context, see Broshi and Barkay 1985; Gibson and Taylor 1994: 25–48; Kelley 2019: 44–60.

Fig. 1: Photograph taken within 24 hours of the excavation (© Photo Elia, courtesy of A. Walls)

However, their finding was only made public in a paired article by C.M. Bennett and Humphreys in 1974.

Bennett 1974; Humphreys 1974.

The stone is situated on a wall that served as part of the foundation for both Hadrian's structure and Constantine's basilica, both of which were backfilled to provide support for the upper constructions.

Kelley 2019: 60.

This location for a detailed drawing and inscription poses some challenges regarding its dating. Two possibilities seem most plausible. M. Broshi and G. Barkay suggest that the drawing and inscription were likely created during the construction of the Constantinian basilica, between approximately 325 and 335 AD.

Broshi and Barkay 1985: 124; see also Kelley 2019: 55.

Alternatively, S. Gibson and J.E. Taylor, considering the archaeological context of the stone and the depiction of the ship, date it to the second century AD.

Gibson and Taylor 1994: 42, 47–48.

Controversy

A comparison of the drawing and inscription, based on the photograph taken within 24 hours of the 1971 excavation (Fig. 1) and another captured in 1975 following a cleaning process, reveals notable differences between the two (Fig. 2).

Fig. 2: Photograph taken after cleaning in 1975 (Gibson & Taylor 1994, 32, Fig. 25)

After the British Institute had left, the Armenian Patriarchate entrusted the Franciscan Fr Emmanuele Testa with the task of cleaning the stone in 1975. The editor of CIIP I.2 787 states that Helms, who discovered the stone with Walls in 1971, baselessly accused Testa of having changed the inscription during the cleaning process.

CIIP I.2 787: the edition is to be found on page 92 of the volume.

However, Helms had valid grounds for raising concerns in this regard.

Outlined in Helms 1980; see also Walls n.d.; Gibson and Taylor 1994: 31–34.

Soon after the discovery of the ship, Helms engaged in a discussion with Father Pierre Benoit of École Biblique in Jerusalem regarding the interpretation of the inscription.

Helms 1980: 109–110.

Following a quick ocular inspection on the excavation site,

Walls n.d.: 4; Helms 1980: 109.

Benoit immediately ended up with a Christian reading domine ivimus, while Helms supported the interpretation Isis Mirionimus. Testa sided with Father Benoit, and it is unclear to us if he was even aware of Helms’ reading. The precise details of the cleaning process remain unknown. According to Testa, the cleaning was performed with a "chemical substance," while other testimonies mention the use of leather and oil, as well as pig fat.

Testa 1976: 219; Gibson and Taylor 1994: 32.

On January 12, 1977, the Criminal Identification Bureau of the Israel Police examined the inscription using infrared illumination. According to Broshi and Barkay, the investigation confirmed that the drawing or the inscription had not been altered in any way.

Broshi and Barkay 1985: 128.

However, as Gibson and Taylor accurately point out, it can be questioned whether infrared illumination can reveal any lines on the lowly porous stone support of this drawing that would not already be visible to the naked eye.

Gibson and Taylor 1994: 33–34.

In any event, the alteration of the original drawing and inscription resulting from the cleaning process appears evident.

See Helms 1980: 106 (fig. 1) and the discussion therein; Gibson and Taylor 1994: 34.

Language

The initial point of interest in the inscription lies in the peculiar verb form ivimus. The editor of CIIP I.2 787 interprets it as the uncontracted first-person plural perfect of the Latin verb ire, meaning "to go," and chooses to translate the phrase as "Lord, we have come" instead of the expected "Lord, we went." This interpretation would presume that the verb ire was also used to indicate arrival. Judging from thousands of occurrences from different periods of Latinity, the verb ire had nearly the same semantic extension in Latin as the English go, denoting every kind of motion for animate or inanimate things: walking, riding, sailing, flying, moving, passing, and so on.

Thesaurus Linguae Latinae, s.v. eo.

In what follows, we investigate whether there is any evidence indicating that ire might also have signified arrival at some stage in the history of the Latin language, in the way it does in modern Ibero-Romance languages. These languages descending from Latin could theoretically continue a previously unknown Vulgar Latin usage, with ivimus attested in the inscription as a unique manifestation of it.

In modern Spanish, the motion verbs ir (< Lat. ire) and venir (< Lat. venire) both express the movement of something or someone from one point to another, using the speaker’s physical position at the moment of the utterance as the point of reference.

Suadoni 2016: 82.

The verb ir indicates speaker’s movement from their current place to another, regardless of the location of the person they are addressing. Consider the following example:

– Esta noche voy a la misa con Dolores. ¿Vienes? (verb venir) "Tonight I’m going to the mass with Dolores. Are you coming?" (verb to come)

– Me parece buena idea: ¡voy! (verb ir) "It looks like a good idea: I’m coming!" (verb to come)

The pattern differs from English and many other European languages, in which the positions of both the speaker and the addressee together determine the selection between the two verbs. While the general behaviour of Latin motion verbs more closely resembles English than Spanish, cases of ire in places where one might expect venire appear in Old Latin, especially in the comedies of Plautus and Terence:

Euge Astaphium eccam it mihi advorsum. "Great! Astaphium comes to meet me." (Plautus, Truculentus 503, c. 200 BC, where it is the third-person singular present indicative of ire)

Unde is? "Where do you come from?" (Plautus, Mostellaria 547, c. 200 BC, where is is the second-person singular present indicative of ire)

However, the system that arises from Old Latin comedies differs from that of Spanish. Venire is consistently the most typical verb in sentences like the ones above, and, most importantly, the question "Where do you come from?" cannot be expressed with ir in Spanish. Rather than indicating the speaker’s position, the Old Latin system seems to be based on Aktionsart: the Latin ire encodes events that are predominantly atelic, i.e., presented as heading for no specific endpoint, while venire encodes events that are predominantly telic, i.e., presented as heading for a specific endpoint.

Ricca 1991: 166–175.

In the present discussion, the crucial point is that due to its very atelicity, ire is hardly precise enough to express the meaning of arrival in an utterance like domine ivimus. The verb is too vague to convey the meaning "Lord, we have come/arrived." Therefore, one would expect domine venimus instead, perhaps accompanied by a clarifying prepositional phrase such as ad te "to you" (cf. mihi advorsum "towards/to me" and unde "wherefrom" above), assuming that the writer pursued the meaning that the proponents of the reading domine ivimus seem to postulate. In summary, surviving Latin sources offer no indication whatsoever to legitimize the reading "we have come" in such brief and uncontextualized phrases as domine ivimus.

Highly pertinent to the present discussion is also the fact that the uncontracted plural perfect ivimus, in contrast to the contracted īmus, is, as far as we can tell, never attested in Latin inscriptions or Classical literature.

Ivimus is attested a few times in Late Latin literature, including one instance in the Vulgate 2 Kings 7,10: ad castra Syriae ivimus ("We went to the camp of the Syrians"). It is worth noting that the Vulgate emerged almost a century after the latest production estimate for the inscription. This detail seems to have eluded scholars who seek parallels between the Vulgate and the inscription in question. In any case, the uncontracted form is characteristic of the Christian writers of Late Antiquity, reflecting their inclination toward morphological explicitness, and certainly was not part of everyday spoken Latin in the Imperial period.

Instead, one might be tempted to interpret ivimus as ibimus, where b has been substituted by v. Contrary to what Gibson and Taylor claim, the spelling v for b is not about "the rare substitution of v for b," but one of the most frequent spelling variations of non-literary Imperial and Late Latin.

Gibson and Taylor 1994: 46.

The original Latin sounds /b/ and /w/, marked with the letters b and v, respectively, had both become a bilabial fricative /β/ in intervocalic position during the first century AD, understandably causing non-classical spellings among the less educated.

Väänänen 1981: 21.

In the absence of specific context, each verb lexeme is bound to appear in its generic meaning. Thus, the perfect ivimus translates neutrally as "we went" and the future ibimus as "we will go." In Imperial and Late Popular Latin, the future acquired an additional hortative meaning, allowing for a translation as "let us go" as well. None of these is compatible with the reading "Lord, we have come" proposed in CIIP I.2 787. The editor supports his interpretation with two references to the Bible which, however, have ire in the future tense and clearly speak of "going," not of "coming" or "arriving:"

The text recalls Psalm 121,1: "In domum Domini ibimus” = "Let us go to the house of the Lord." There could also be an allusion to Jn 6,68: "Domine ad quem ibimus?" = "Lord, to whom shall we go (turn)?" It expresses the relief felt by Christian pilgrims arriving in the Holy Land and being close to Christ’s burial place.

CIIP I.2 787: 93.

Gibson and Taylor also appear to confuse going with coming. While discussing the interpretation "Lord, we will go," they first duly note that:

a Future tense is not consistent with an image that shows a completed action. The juxtaposition of the drawing of a ship in port after its voyage and the inscription which would naturally be in the Perfect tense is logical.

Gibson and Taylor 1994: 41, 46; the authors' discussion makes it clear that the action is completed, as the drawing depicts a ship in harbor with a lowered main mast.

After that, however, they hypothesise about an alternative reading, domino ivimus, and translate it as "we came to the Lord."

Gibson and Taylor 1994: 47.

This seemingly unconscious mistake is understandable because "to come" is, indeed, the verb one expects to see in a context of arriving somewhere, not "to go." The generic Latin verbs that denote coming to or arriving at a place are venire and advenire or, in Later Latin, also plicare (se) (originally "to fold"), jungere se (originally "to join"), and arripare (originally "to go ashore").

Väänänen 1981: 21.

Broshi is more systematic and translates "Lord, we went," while referring to Psalm 121 [122], which he describes as the classical psalm of the pilgrim to Jerusalem.

Broshi 1977: 349.

He interprets domine ivimus as "the joyous exclamation of pilgrims who sailed from the western part of the Empire and finally reached the Holy City." Broshi and Barkay, however, opt for the translation "Lord, we shall go," duly mentioning the substitution of b with v and associating the phrase with the Psalm 121 [122] and Jn 6,68.

Broshi and Barkay 1985: 125; cf. §14 above.

The difficulty is that such parallels can be drawn almost endlessly. A convincing parallel between an inscription and a biblical passage requires unambiguity. However, this clearly does not apply to the phrase domine ivimus, as the first word is one of the most common in the Vulgate, and the second a generic verb. We also find it puzzling why a Latin-speaking pilgrim, who has just arrived in Jerusalem, would express "his relief" or "joy" in the future tense, even if such constructions can be found in random psalms. One might wonder why the pilgrim did not choose phrases like Ezr 8,32: venimus adorare eum ("we have come to worship him") or Mat 2,2: venimus Hierusalem ("we came to Jerusalem").

Paleography

While Gibson and Taylor conclude that there are significant alterations in the current drawing compared to the original, they nonetheless choose to support the domine ivimus reading of the inscription, although doubting its Christian content.

Gibson and Taylor 1994: 33–34.

Through an analysis of the 1971 photographs (reconstruction in Fig. 3) and certain paleographical considerations, they reach the conclusion that the incontestable letters are D-MIN--VIMUS (with three uncertain letters) (Fig. 3).

Gibson and Taylor 1994: 42–45.

Fig. 3: Vestiges of the inscription evident from the analysis of the 1971 photographs (Gibson & Taylor 1994, 45, Fig. 34)

We find this perplexing because their reconstruction could easily be interpreted as Isis Mirionimus as well. The first two letters are clearly visible in Fig. 1: it simply reads IS and not D, followed by I and S. This reading is also supported by the fact that there is a discernible gap after the first four letters (or two if one opts for the reading DO-). The next letter is certainly M, followed by a somewhat uncertain R and I. We cannot read any letter between M and R. I is followed by an O, lacking part of its left side. The last four letters, -IMUS, are indisputable.

So, the question arises: Why do Gibson and Taylor insist on the domine ivimus reading? The reason appears to lie in a linguistic mistake or what they refer to as a "gender problem."

Gibson and Taylor 1994: 44.

Quoting Humphreys, Taylor and Gibson correctly mention that "ISIS MIRIONIMUS is a phonetic transcription in Roman letters of the Greek ΙΣΙΣ ΜΥΡΙΟΝΥΜΟΣ."

sc. ΜΥΡΙΩΝΥΜΟΣ.

Next, however, they assert that "Isis, in fact, is never qualified by the masculine -os ending," citing CIL III 882 (Isidi Myrionimae) in support of their argument. In reality, the common Greek epithet qualifying Isis, μυριώνυμος, is not masculine but feminine.

For a general discussion of the epithet, see Bricault 1994: 67–86.

Both the masculine and feminine forms of compound adjectives, like μυριώνυμος, exhibit identical endings to those of 2nd-declension masculine nouns in -ος. Furthermore, contrary to what Gibson and Taylor claim, Isis is in fact qualified by the masculine-like ending in Latin inscriptions, as exemplified by Isidi Myrionymo (AE 1956, 244), which directly renders the Greek dedication Ἴσιδι μυριωνύμῳ in Latin script.

Also attested as Isidi Mirionymo (AE 1968, 230) and Isidi Myrionimo (Grisonic et al. 2022: 234) with ‘i’ instead of ‘y’.

We are tempted to assume that Gibson and Taylor would have been more inclined to support the Isiac reading had they been aware of this fact.

The Ship

The drawing of the ship depicts a typical Roman merchant vessel, known in Latin as navis oneraria.

Casson 1971: 169, 175 (third type); Gibson and Taylor 1994: 35.

Isis was strongly associated with the sea.

See Bricault 2019.

As protectors of navigation, Isis and her consort Sarapis are known to have lent their names to ships.

See Bricault 2006: 168–170 and the commentary on RICIS 115/0401.

A case in point is the grain vessel depicted in a third-century AD Ostian fresco named Isis Giminiana.

CIL XIV 2028 = RICIS 503/1132; see also Helms 1980: 105; Nagel 2019: 1110.

Another example is a trireme named Isis, attested in several first- or second-century AD epitaphs.

RICIS 501/0218 = CIL VI 3123; RICIS 504/0501 = CIL X 3615; 504/0502 = CIL X 3618; RICIS 504/0503 = CIL X 3640. For more examples, see the commentary on RICIS 115/0401 and Nagel 2019: 703 f.

Hence, it is entirely plausible that a ship could bear the name Isis Mirionimus.

Conclusion

In this paper, we have shown that the reading domine ivimus is improbable, given the meaning of the verb ire and the apparent absence of the uncontracted form ivimus in both spoken language and most literary Latin texts. We consider the reading Isis Mirionimus to be the most likely, although acknowledging the need for some doubt as the original inscription seems to be lost. We cannot but endorse Gibson and Taylor’s call for the Armenian Patriarchate to commission an independent analysis of the pigments, which we expect will reveal the tampered parts of the inscription more clearly than is possible with visual inspection alone.

Gibson and Taylor 1994: 34.

Bennett, C.M. (1974) "The Jerusalem Ship," IJNAUE 3 (2): 307–309. Bricault, L. (1994) "Isis myrionyme," in C. Berger, G. Clerc & N. Grimal (eds), Hommages à Jean Leclant 3. Études isiaques. Le Caire. Bricault, L. (2006) Isis, Dame des flots. Liège. Bricault, L. (2019) Isis Pelagia: Images, Names and Cults of a Goddess of the Seas. Brill. Broshi, M. (1977) "The Jerusalem Ship Reconsidered," IJNAUE 6 (4) (1977), 349–356. Broshi, M. and Barkay, G. (1985) "Excavations in the Chapel of St. Vartan in the Holy Sepulchre," Israel Exploration Journal 35 (2/3): 108–128. Casson, L. (1971) Ships and Seamanship in the Ancient World. New Jersey. Gibson, S. and Taylor, J.E. (1994) Beneath the Church of the Holy Sepulchre, Jerusalem. The Archaeology and Early History of Traditional Golgotha. London. Grisonic, M., Cesarik, N., Vilogorac Brčić, I. and Štrmelj, D. (2022) "Calpurnia L. Pisonis filia, Cn. Pisonis neptis and the Votive Altar Dedicated to Isis, Serapis, Osiris and Anubis in Caska Cove, on the Island of Pag," Vjesnik Arheološkog muzeja u Zagrebu 55 (2): 231–255. Helms, S. (1980) "The Jerusalem Ship, ISIS MYRIONYMOS, and the True Cross," IJNAUE 9 (2): 105–120. Humphreys, S.C. (1974) "The Jerusalem Ship – Analysis and Dating," IJNAUE 3 (2): 309–310. Kelley, J.L. (2019) The Church of the Holy Sepulchre in Text and Archaeology. Oxford. Nagel, S. (2019) Isis im Römischen Reich. Teil 1: Die Göttin im griechisch-römischen Ägypten, Teil 2: Adaption(en) des Kultes im Westen, Philippika, Altertumswissenschaftliche Abhandlungen (Contributions to the Study of Ancient World Cultures 109). Harrassowitz. Ricca, D. (1991) "Andare e venire nelle lingue romanze e germaniche: dall’Aktionsart alla deissi," Archivio Glottologico Italiano 76: 159–192. Suadoni, A. (2016) Verbos de movimiento, deixis y proyección metafórica: el caso de andare y venire en contraste con ir y venir (Tesis doctoral, Universidad de Granada). Testa, E. (1976) "Il Golgota, Porta delle Quiete," Studia Hierosolymitana in honore di P. B. Bagatti. Jerusalem. Thesaurus Linguae Latinae. Leipzig/Berlin, 1900–. Walls, A. (n.d.) Church of the Holy Sepulchre 1971 Excavations Clarified: "The Jerusalem Ship" — Interpretation and Lies (https://www.academia.edu/44876770) Väänänen, V. (1981) Introduction au latin vulgaire. Paris.