The Iconography of Children as Cultic Characters in Mediterranean tophet Precincts

I Talking monuments and levels of communication

In light of the above, stelae from tophet precincts might be designated talking monuments with different levels of communication.^[10] Effectively, they are specific and peculiar to each sanctuary (in materials, in typologies of the blocks, in iconography and in the style of their depictions), and contribute to the creation of a local religious landscape that represents a sort of identity card for the corresponding settlement.^[11] The craft and production of these stelae derived, at least in part, from the long metropolitan Levantine tradition of ritual stone-markers known as baetyl (or “house of the god”).^[12] Yet, despite of these shared origins and certain basic trends emerging throughout the central Mediterranean, each sanctuary shows its own unique repertoire characterized by distinctive elements which allows us to study its particularities and originality.^[13]

Concerning levels of communication, a stela conveys various meanings which are transmitted by two specific languages, the figurative and the written, only hypothetically related to each other. The figurative message, or iconography, consisted of images and symbols of various types, elements of a code that we can attempt to decipher only from our external point of view (etic perspective). Although in antiquity the perception and understanding of these depictions must have been immediate and intuitive, modern interpretations of such figurative messages diverge from those of antiquity. The written message, instead, communicates directly to the reader and, merely by its presence, conveys a powerful signal even to those who could not read, probably the vast majority, conferring upon that ritual action an aura of solemnity.

Images may represent a certain realism in the sense that iconography may create a reality in and of itself. Nevertheless, one must not underestimate the power of visual strategies in elaborating and transmitting specific messages, which go beyond simple realistic representations. In any case, the textual sources and ancient figurative motifs should be analysed in a complementary way.^[14]

In the case of Phoenician iconographic repertoires, one cannot hope to infer the specific criteria that regulated the choice of motifs and the connections between the various figurative elements. Despite a limited repertoire of images generically linked to rituals conducted in the precinct, it remains difficult to determine the meaning behind figurative motifs, symbols, and iconographic change through time. At most, one can assume a sort of generic code or repertoire of sacred motifs that ancient communities used in ritual contexts.^[15]

Even in the few cases where text and image are juxtaposed, we remain ignorant of the selection criteria. The inscriptions provide little help in understanding the associated iconography, which, beyond a superficial reading, resists deeper interpretation. In other words, it seems that text and images proceed in parallel, without engaging each other. From a purely archaeological point of view, the general lack of reliable stratigraphic evidence does not allow us to track precise chronological sequences or the evolution in representations over time. Several interpretations have been proposed based on stylistic changes, but only as conjectures.^[16] Given the limits on our knowledge of Phoenician religion,^[17] a full understanding of iconographic motifs remains a desideratum.

Despite such interpretative problems, intensive excavations of tophet precincts across the last century has led to the publication of thousands of stelae from several sites – Carthage and Hadrumetum in north Africa; Nora, Sulci, Tharros, and Monte Sirai on Sardinia; Motya on Sicily, producing a vast corpus of figurative repertoires that relate, directly or indirectly, to various aspects of votive ceremonies.^[18] Due to the great number of finds and their heterogeneous range of motifs and styles, sculpted votive stelae are among the most prominent Phoenician-Punic works of art. Unfortunately, narrow focus on their artisanship has diverted attention away from a deeper evaluation of their function as secondary features in contextual association with the primary archaeological features (i. e., the urns).

From the 1950 s to the 1990 s, scholars applied different methodologies to the study of the iconographic and stylistic features of stone markers. The main advances focused on isolated motifs (particularly in the analyses of Hours-Miédan, C. Picard, Moscati, and Brown).^[19] As a result, we now have an in-depth understanding of the main typological and stylistic traits of this category of monuments and a series of catalogues based on formal analysis (using stylistic or aesthetic criteria), but few rigorous attempts have been made at level of the functional analysis (i. e., using contextual criteria by reconstructing stratigraphic contexts and relations to urns). Priority should be given to the study of iconographies attested at specific sites during specific chronological phases – figures/characters (primary or secondary, human or faunal), tools of ritual significance, astral and plant motifs, etc. – however, no attempt has been made to systematically establish links to cultic life, thus much work remains, especially at the level of historical synthesis.

The anthropomorphic representations of human beings and symbolic representations of abstract subjects both appear on votive markers and both have been related to the divine sphere, with the figural representation of the divine categorized as iconism and the abstract as aniconism.^[20] While both types of imagery have been taken to represent deities,^[21] abstract aniconic imagery has almost universally been interpreted as an abstraction of the divine; yet we lack clear proof that they refer to divine entities. Therefore, it seems prudent to consider the aniconic as generic allusions to the ritual/cultic contexts but nothing more specific. The predominance of one category (figural/iconic) or another (abstract/aniconic) depended upon the specific settlement and the historical period – each sanctuary had a predilection for specific iconographic elements, each preferring its own repertoire and each witnessing stylistic variation over time.^[22] For example, at Sulci, local production emphasized anthropomorphic characters, depicting both female or male figures in various postures (e. g., typical motif of the so-called standing woman holding a disk to her breast).^[23] Carthage and Nora, instead, showed a preference for abstract symbols, e. g., the baetyl (single or multiple), lozenge, “Sign of Tanit,” solar disc and lunar crescent, “bottle idol,” etc. (see Fig. 1).^[24]

Fig. 1

limestone stela from Carthage (G.-C. Picard 1955, Cb – 368) – symbol of lunar crescent above solar disc, “Sign of Tanit,” and “bottle-idol” on pedestal (G. R. Swain photograph 5.680, from 1925 F. W. Kelsey Carthage Excavation Archive, used with permission ©2023 Kelsey Museum of Archaeology, University of Michigan).

II New perspectives on anthropomorphic depictions of children

The abstract representations of the divine, ranging from the so-called “Sign of Tanit”(plausibly derived from the Egyptian ankh), to the blessing hand, caduceus, solar disc and lunar crescent, etc., all convey a symbolism too cryptic and too general to provide precise information about ritual activity, with the possible exception of the “bottle-idol” (as discussed below).

Leaving aside for now the iconographic repertoires in general, I draw attention here to specific anthropomorphic depictions, particularly those of children, which can rarely be identified with certainty as figures on cippi and stelae. In this regard, two questions arise: 1) Why are explicit representations of infants so rare, given the relevance of the very young children buried below these stone markers? 2) Are there any discernible symbolic allusions to children as either dead or dying? One can occasionally distinguish two different patterns on stelae: one simple, depicting a child alone; the other composite, representing a child together with an adult. The tophet precincts of Carthage and Motya provide examples of the former (simple) pattern, while the latter (composite) pattern occurs on Sardinia, especially at Sulci, Tharros, and Monte Sirai. The Carthaginian iconographic corpus includes various human figures, nevertheless they are represented in such an extremely simple and rudimentary way that they cannot be interpreted with certainty, as either adults or as children.^[25]

In some rare and chronologically disparate examples, however, the representation of a child is undeniable, as is the case of one stela from Carthage, as well as others from Motya. The Carthaginian (fifth-fourth centuries) has a swaddled infant, in frontal view, set below a lunar crescent and dominating the scene.^[26] The highly detailed face deserves special attention, although its traits more closely resemble an adult than an infant.

The examples from Motya (second half of the sixth-fifth century) depict infants/new-borns in profile, either naked or wearing diapers.^[27] These characters are faceless, in contrast to the detailed representation on the Carthaginian stela. The data at our disposal are too limited to permit precise assessment of chronology and style, but a reasonable supposition would attribute differences to local figurative tendencies and/or variation over time.

As far as the second pattern (composite), the so-called “priest-stela” from Carthage (Bardo Cb – 229) provides the most famous example, already the subject of countless publications and analyses (Fig. 2a).^[28] The adult male figure raises his right hand in the traditional blessing posture and holds in his left arm a child, who is evidently alive. Another comparable case of this iconographical pattern comes from the sanctuary of Motya (Fig. 2b), which presents an adult in an identical posture, carrying something (or someone) in his/her left arm.^[29] Despite the chronological and geographical distance between these two examples, dating back from the sixth-fifth centuries (Motya) to the fourth-third centuries (Carthage), the occurrence of the same iconographical pattern across time and space seems a noteworthy trend towards the preservation of traditional motifs over centuries.

Certain composite representations that pair an adult and child come from Sulci, Monte Sirai and Tharros, all on Sardinia and all dating to the fourth-second centuries. Those from Sulci and Monte Sirai show two figures in frontal view: the former from Sulci (fourth century) presents the child held in front of the (presumably male) adult;^[30] of the latter from Monte Sirai (first century), the two figures are set besides one another and are depicted in a more schematic way.^[31] The gender of the adult is unclear (female?) but, significantly, the child is again held in the left arm.

The example from Sulci evokes an emotional response, and the scholarly interpretation of the scene has a peculiar history. Initially, Moscati considered it as a binomial depiction (adult/behind + child/in front), but he later reinterpreted the scene as a “standing woman holding a disk”, one of the traditional motifs of the Sulci repertoire.^[32] Moscati attributed the strangeness of this iconography to confusion on the part of the artisan, who had misinterpreted the common leitmotif of the standing woman. On closer inspection, Moscati’s re-interpretation lacks support. No objective reason accounts for the shift away from his first interpretation, since the scene clearly reproduces two human figures, an adult and a child. This rare motif likely held special meaning for the artisan, based both on artistic tradition and religious custom, conforming to standard practices of the sanctuary. Since the tophet represented the social life of the community and since the ritual undertaken in that precinct would have had by a consumer base, the carving of the stela would not have been performed by an artisan unaware of their intended market. Indeed, even if our knowledge of engraving techniques is limited, I presume that the preparation of the monument required several working days, if not weeks. Over that period, any misunderstanding or error by the artisan would have been noticed and corrected, or the stela rejected, if flawed in its execution or different from the wishes of the client. On the contrary, in the case of this “misunderstood” stela from Sulci, it had to have been accepted by the client and it had to have been installed in the sanctuary, fulfilling its functions as an ex-voto. These considerations lead me to reject Moscati’s second hypothesis, and to interpret such a scene as representing a ritual act, i. e., one precise moment of the ceremony related to consecrating a child as a votive offering in front of an adult figure (apparently male, possibly a priest or another ritual officiant).^[33]

Fig. 2a

limestone “priest-stela” from Carthage – adult and child (Icard C217, in Hours Miédan 1950–1951, pl. XXXV; G.-C. Picard 1955: Cb – 229).

Fig. 2b: sandstone cippus from Motya – adult and child (based on Moscati and Uberti 1981: Tav. CLXXI, no. 951).

The sanctuary of Tharros also offers some interesting composite representations of paired figures, e. g., stela no. 142 (fourth century) depicts two characters – an adult and a child.^[34] The two figures are in profile, facing each other, and incomplete, but again note the composition – adult to the viewer’s left, child to their right – and note the particular head covering of the adult, typical of a ritual officiant.

III Iconography and epigraphy

While the few cases studies offered in this brief survey could be expanded, we have seen enough to draw some preliminary conclusions. First, one must acknowledge the general anonymity of children at the figurative level in the iconographic repertoire of the tophet, as reflected both in the rarity of infant representations and the rarity of their epigraphic documentation. Inscriptions allude only infrequently to children as the object of sacrifice (with the expression š^ↄry, mentioned below, is an exception), yet even as the objects/subjects of the ritual, they always remained unnamed.^[35]

Given the central role that children played in the tophet precincts, this peculiar phenomenon can be perhaps understood on a comparative ethno-anthropological basis. In many ancient cultures, new-born babies had no identity, no name. An infant was anonymous at its birth and did not receive a name before being integrated into society through ceremonies that varied from one culture to another and from time to time.^[36] A different attitude on the part of the living pertained to this specific category of the deceased, an attitude partly manifest in ritual practices. An infant was not yet a full member of society, therefore it could not be considered in the same way as an ordinary (adult) decedent. Consequently, we may suppose that for this reason the death of a child – a new-born – tends to be archaeologically invisible in contexts related to funerary ceremonies.^[37]

Therefore, we assume that the individual child was not directly remembered in the tophet, even figuratively, except for those few anthropomorphic depictions noted above. At the abstract level, a key symbolic representation is the motif of the “bottle-idol,” one of the most common and prolific symbols found in the western Phoenician world, with variations across a chronological range and geographical span from Carthage (Figs. 3a–b) to Sousse and from north Africa to other Mediterranean tophet precincts in Sicily and Sardinia (save Sulci, where this motif has not yet been attested).

Fig. 3a

limestone stela from Carthage – lunar crescent above “bottle-idol,” with inscription CISem. I.3231 (Louvre AO 5182).

Fig. 3b: limestone stela from Carthage – “bottle-idol,” with inscription CISem. I.1532 (Louvre AO 23410).

Despite regional and chronological variations, the motif typically depicts a flask with a low neck, cylindrical or round shouldered, sometimes ovoid, with a flat base, resting upon a support (possibly an altar). Several hypotheses have been advanced as to the symbol’s derivation and meaning, as follows:^[38]

Regardless of the debates over the interpretation of the sign, an important transition must be noted. After a period in which anthropomorphism was absent or latent, after a certain period (approximately since about the fourth century), the “bottle” – particularly its neck – undergoes a process of explicit anthropomorphism, gaining a human face, provided with eyes showing various expressions, and sometimes feet. From abstract inanimate object, the bottle comes to life and begins to represent a living character, bearing some similarities to a child.^[39]

Here we might draw the following inferences. 1) Regarding the “bottle-idol,” irrespective of its original associations, one can recognize a fusion of iconographic motifs, combining a representation of the sacrificial urn and a representation of a human infant. This fused symbol takes on progressively new significance in western Phoenician contexts, with a transition from an initially abstract, schematized representation to an anthropomorphic one (similar to the transition from aniconic/abstract to iconic/anthropomorphic images of the divine). 2) The varied forms of anthropomorphism,^[40] can be interpreted as an emerging desire to make explicit the main protagonists of the tophet and its rituals. The character of the infant, sometimes swaddled and sometimes in diapers, had been represented in the abstract by the community, since it lacked an individual identity. 3) Considering the specific chronological phase during which this took place, it seems evident that such a process developed in a peculiar social and emotional context. The last phase of Carthage’s existence was a time of extended conflict and high tension, both psychological and social. Hence I would hypothesize that the tradition of abstract or humanization /symbolic representation was broken at that time in favor of a more explicit communication, no longer symbolic and intuitive but now clear and direct.

In conclusion, let us take note of one relationship between iconography and text, precisely concerning the “bottle-idol.” The Carthaginian stela CISem. I.5689 (Fig. 4) has the form of the “sign of Tanit” with a “bottle-idol” lightly engraved in its upper rounded part. The “bottle-idol” image is anthropomorphized/humanized in a way that suggests an infant:

This inscription designates the dedicant as a scribe, belonging to a family of scribes, an élite administrative, political and religious status. The text uses the verb nš^ↄ (“to raise up”) instead of usual ndr (“to vow” or “to devote”), and ends with the terms ^ↄyt ^ↄršt šry, marking the object dedicated “as the chosen offspring” (literally “as the chosen/requested [part] of his flesh”). It provides one of the rare direct epigraphic mentions of a child as the object offered to the gods, combined with a symbolic allusion to the humanity of the victim, expressed by an anthropomorphized form of the bottle-idol. Preliminary research into Carthaginian stelae using terms that may denote human victims – e. g., with the object marked as mlk (“offering,” literally “a causing-to-go-forth”) or similar expressions – have provided further cases of association with an anthropomorphic bottle-idol.^[41]

The proposals in this paper represent an attempt of reading and interpreting some iconographies of the tophet. Finding more explicit references to ceremonial activity in Phoenician world may prove challenging, since the main problem in this objective remains the lack of archaeological evidence and reliable connections to stratigraphic contexts. However, a careful and more systematic analysis of the various repertoires from different Phoenician and Punic settlements of the west would be worth the effort, in order to contribute to expand the range of the documentation at our disposal.

Fig. 4

limestone stela from Carthage in the form of the “sign of Tanit” – “bottle-idol” symbol with inscription CISem. I.5689 (Picard 1955: Cb – 552).