Visual Trope and the Portland Vase Frieze: A New Reading and Exegesis

Arion Vol. 2.1


Among the extant masterworks of Roman art, there is probably none that has generated more scholarly debate than the Portland Vase, especially over the interpretation of its elegant frieze. Since the discovery of the vase about four centuries ago, no fewer than forty-four different theories attempting to interpret the scenes on the vase have appeared.1 In the main, the theories fall into two categories, those relating the frieze to Greek myth, and those linking the figures to Roman personages. Moreover, there is no consensus whether the frieze is two separate scenes or one continuous story. D. E. L. Haynes, who believes the frieze depicts the story of Peleus and Thetis in one scene, holds the high ground in the Greek myth camp. E. Simon leads the opposition with her theory that the frieze is two scenes depicting a bit of Julian propaganda chronicled by Suetonius — that Augustus was the product of his mother’s congress with Apollo in serpent form.2 Both theories are founded almost exclusively on the authors’ interpretation of the serpent in Figure C (see illustration).

The obstacle to interpreting the frieze is that all but one of the figures lack identifying attributes. Figure B is clearly Eros; the others offer no ready clues to their identities. D.B. Harden succinctly summarizes the situation: “The vigour and ingenuity with which so many scholars have debated the problem demonstrates nothing so much as that the subtle distinctions on which most of the identifications are based are not a firm foundation for interpretation in our present state of knowledge.”3

Given the Augustan dating of the vase, some degree of ambiguity is understandable. As P. Zanker notes in his discussion of “Atticizing” compositions such as the Portland Vase, “Evoking a whole field of associations was more important to the artist than iconographic precision.”4 However, the dearth of attributes in the Portland Vase frieze, surely intentional, prompts the viewer to seek some other method by which the scenes can be read. Considering the probable value of the cameo glass vase and the time necessary to execute such a work, it is unlikely that the frieze has no meaning and is purely decorative.5 It may be assumed, then, that there is meaning in the frieze and that there is some way to read it.

In a discussion of private imperial art in the age of Augustus, Zanker suggests that the frieze of the Portland Vase is “a kind of erudite puzzle [and that] a deliberate ambiguity must have been contrived by the artist as part of the image’s appeal.”6 A parallel between the nuance and innuendo of contemporaneous Roman poetry and the artistic understatement of the frieze suggests itself.7 In this case, the puzzling frieze might be reflecting a puzzling poem, one in which epithets — the literary counterpart of attributes — were used sparingly. As R. Brilliant points out, “The parallel between literature and the visual arts has been a philosophical and critical topos since antiquity.”8 The Roman elite, those who would be unraveling the “erudite puzzle” on this exquisite piece of private art, were undoubtedly familiar with a poem noted for both its elliptical nature and its paucity of epithets: Catullus 64, “The Marriage of Peleus and Thetis.”9 Could the innovative work of Catullus have inspired an equally innovative work of art?

I propose (1) that in lieu of standard iconography, a synthesis of visual and literary tropes in the form of hidden letters, that is, cryptograms, tells the viewer how to read the frieze; (2) that scene A of the frieze illustrates specifically lines 26-30 of Catullus 64 by visually rendering the epithets and tropes used in those lines; (3) that visual reproductions of other tropes in Catullus 64 further substantiate the relationship between frieze and poem; (4) that scene B of the frieze is an intentional aberration from the poem; (5) that there is a complex and unfinished interaction between scenes A and B which the viewer must complete; and (6) that in the final analysis, the frieze duplicates both the theme and central structure of Catullus 64.10

Looking upon the frieze as a whole, one is immediately struck by the balance of composition in its two pendant scenes. Both are flanked on the left by a young man facing right and on the right by an older figure facing left; a female figure reclines in the center of each. Both scenes begin with a piece of architecture; both contain rocky piles and a three-limbed tree.

With these similiarities drawn, one must then decide on which side of the vase to begin an interpretation, scene A or scene B. Where to begin is, of course, crucial, yet previous interpretations seem to disregard certain visual clues in the frieze and arbitrarily designate either one scene or the other as the first.11 Several aspects of the composition convince me that scene A is undoubtedly the first of the two: (1) In addition to the fact that Eros, figure B, is the first who can be firmly identified, he is also the only figure without a pendant. Because he disrupts an otherwise perfect balance in the composition of the pendant scenes, as well as provides the only explicit visual clue to an interpretation, it seems clear that the eye of the viewer is purposefully being drawn to scene A. (2) The architecture which begins scene A is positioned at the edge of the frieze; that of scene B is set noticeably back from the edge. Because one reads the frieze from left to right — the direction of motion of figures A and B — the flow of the ground line is uninterrupted when one begins with scene A. If, however, one begins with scene B, the architecture of scene A acts as a definite impasse. (3) The Pan-like heads which separate the scenes show distinct differences in mood by their facial expressions and yet prior interpretations take no account of this. The head preceding scene B has its face set in a definite grimace. The one preceding scene A, however, displays a smile; the arrangement of the beard reinforces the mood in each case. It would seem, therefore, that the viewer is being urged to begin where the mood is most auspicious. Hence my interpretation begins with scene A.

The presence of Eros suggests that love is the most viable theme. The fact that Eros carries only a bow and no arrow — implying that the love dart has already been spent — and that the lead position in each scene is held by a virile young man, reinforces the idea that the scenes depict courtship in progress.

The next most noticeable aspect of scene A is that the eyes of figures B, C, and D are all focused on figure A. Should we not focus our attention upon him as well? But again, the figure seems to possess no identifying attributes and our reading seems blocked, unless, perhaps, the placement of the figure can offer some clue. Careful scrutiny of the architecture against which the figure is placed reveals several peculiarities. First of all, its two piers are markedly mismatched: the one on the left is in a bold three-quarters view, while the one on the right, distant by only half the length of figure A’s arm, is wider, full frontal and flat. It is, in fact, the only two-dimensional element in the entire frieze. These dissimilar verticals support a Doric entablature upon which Ionic dentils are carved, a combination of two orders which, according to the Augustan architect and author Vitruvius, “will jar upon us.”12 The number and placement of the triglyphs conflict with architectural norms.13 When compared to the ordered balance of the other vegetation in the frieze, a shrub juts rather wildly out of the left-hand pier, almost as if it had sprung directly from it. Note, too, that like its pendant, the lone pier in scene B also has an odd growth to its left: a truncated, finger-shaped limb pointing directly at it. There can be little doubt but that the viewer’s attention is being drawn to the architecture. The peculiarities of the architecture in scene B strengthen the conclusion. Although the single pier there stands alone, its capital displays clamp holes, that is, points of attachment to other pieces of architecture, as if it were disassembled from some other structure or waiting to be joined to it. There is a detached column base which also shows a clamp hole lying askew at the feet of figure F. All who have studied the Portland Vase acknowledge that considerable thought and attention to detail went into the planning of the frieze, yet no one has offered an explanation for these clamp holes. They are surely significant. There are only three basic elements in the frieze: figures, landscape, and architecture. Since the figures, except Eros, are unidentifiable in and of themselves (as four hundred years of study and forty-four theories so prove), and the landscape features serve to draw attention to the aberrant architecture, it seems very likely that understanding the role of the architecture will provide the key to unlocking the meaning of the frieze.

If we begin with the first feature of scene A, the two oddly mismatched piers, and name them as a Roman would, in Latin, pilae, then the accented sound of the first syllable of pilae, pi, leads us to recognize an oblique visual metaphor: the Greek letter pi, artfully dissimulated in the Greek architecture. Turning then to the other piece of architecture, we immediately see an iota. Although there is yet no other known example of this phenomenon in Roman art, the rationale behind these architectural cryptograms is not difficult to trace. One needs only to consider two factors: (1) The style of Roman wall painting contemporaneous with the Portland Vase is often labeled “architectural fantasy.” Many examples of eccentric architectural forms dated to this period testify to the imagination and creativity of the artist. In fact, fantastic architectural creations on Augustan walls reached such extremes and enjoyed such popularity that Vitruvius came to lament the style.14 So it is easy to imagine that the creation and camouflage of Greek letters within Greek architectural forms would be in accord with the artistic taste of the day. Furthermore, disassembled letters were by no means uncommon at this time; both Caesar and Augustus used them in ciphers.15 (2) To have architectural features creating an acronym which spells out the very word that pronounces the first letter, is a visual example of the kind of playfulness with words that was popular among the ancient Romans.16 I call it a visual trope. In this case I am not only talking about “trope” in the sense of using words figuratively — as when their accustomed range of meanings is suddenly expanded — but in the sense of using architectural figures as words. The supportive structures are no longer merely background features of the landscape; it is apparent that they are serving a double role.

It does seem likely that the letter pi and the word pi are, somehow, the keys to the puzzle. Let us first consider the letter pi. As noted above, the two verticals that support the entablature are quite mismatched; they differ in view, color, and dimensionality. Such noticeable mismatching urges the viewer not to pair the right vertical with the left but simply to see it for exactly what it appears to be: a pilaster. By assuming that the right-hand vertical is a pilaster, a nonsupporting, decorative element, two questions are raised: (1) How does it signify? (2) How is the entablature supported? Let us note that figure A obscures the greater part of the pilaster; that he is visually much the stronger of the two elements, and that the entablature appears to rest upon his head. Figure A, then, may easily be construed as a telamon. Perhaps then the pilaster, parastas in Greek, serves only to form the right vertical stroke of the pi, a role to which it is ideally suited by its initial letter, pi, and, as the backdrop of the superimposed figure A, to suggest his role as telamon. The latter function of the pilaster then also plays upon the Doric style of architecture; Vitruvius twice likens Doric columns to naked men.17 Therefore, if we construe figure A as a telamon — a member of the architecture — he must also be the right-hand stroke of the letter pi. It now seems a matter of simple logic to assume that the name of figure A begins with the letter pi.

Besides the letter pi, the word pi also holds a trope. Not only does it denote the letter pi, but it is homonymic with an adverb in the Doric dialect meaning “Where?”, pi epsilon iota, which, in fact, is also an earlier spelling of the word for pi, and — transliterated — of the first syllable in pila. Bearing in mind that the architecture is Doric, and that Doric is the dialect of many lyric poets (there is no Doric literary prose), the trope may be asking actually two questions: “Where are these scenes taking place?” “Where in poetry do these scenes occur?” Apparently, when the complex question “Where?” is answered, the viewer will be able to read the narrative. Given the acrostic structure and meaning of the trope, it is reasonable to assume that the cryptograms supply the initial letters of the place names for each scene. Scene A must be a place whose name begins with pi; scene B with iota.

The architectural tropes have provided sufficient clues to answer the two questions they themselves pose. The pi cryptogram with figure A is, I believe, a unique and precise visual translation of the epithet that the mythological hero Peleus bears in Catullus 64, line 26: Thessaliae columen Peleu. “Peleus, the Pillar of Thessaly” and, consequently, a direct link to the poem. This trompe l’oeil reproduction of Peleus’ two-part epithet seems unmistakable, for just as the words columen Peleu syntactically interact with the genitive Thessaliae in the epithet, so Peleus, the “Pillar,” interacts with the pi cryptogram, which, like its counterpart Thessaliae, names a place: Pelion, the mountain in Thessaly where the Peleus and Thetis match was arranged. Figure A, therefore, I identify as Peleus, and the location of scene A as Mount Pelion. Excellent cause exists to suspect that the other figures may also be identified from the poem, and, indeed, lines 26-30 of Catullus 64 mirror scene A of the frieze:

And now I address you Peleus, O Pillar of Thessaly,

To whom the bridal-torches bring such rich blessings,

To whom the father of the gods, Jupiter himself, has yielded a love of his own.

Did fairest Thetis, daughter of Nereus, capture your heart?

Did Tethys and Oceanus, who circles the world with a river of sea, give you their grandchild to wed?18

Eros, eyes fixed on Peleus and bridal-torch aloft, leads him to his bride-to-be. Zeus (Jupiter), figure D, young as he would be in the Golden Age of mythological time, looks on in approval of the betrothal. This depiction of Zeus is somewhat atypical of his usual iconography; many previous theorists, in fact, identify figure D as Poseidon based on subtle traces of his iconography they rightly detect in the figure.19 The anomaly is easily resolved when one reflects upon the mythology of Peleus and Thetis. Both Zeus and Poseidon, the mythographers tell us, courted Thetis but abandoned their suits when informed of a prophecy that the son of Thetis would be greater than his father; together, they then arranged the Peleus-Thetis match. Figure D, therefore, incorporates subtle attributes of both Zeus and Poseidon in order to reflect this aspect of the myth, and, perhaps, to provide a “red herring.” The intended result, it seems, was achieved, for although the figure evokes both Zeus and Poseidon, a sure identification is possible only by the intended means: recognition of the visual tropes and familiarity with Catullus’ poem.

Figure C is also quite remarkable, for it is the second trompe l’oeil translation of an epithet from Catullus 64 and contains both grandparents of Thetis precisely as depicted in the text. Catullus names Oceanus in line 30, where he uses a literary trick “to bring together in a single context attributes not rationally reconcilable . . . Oceanus, the grandfather of Thetis, and Oceanus, the river.”20 The artist has again duplicated Catullus’ textual trope with a visual one: we see Oceanus in his sea-serpent form to symbolize the “river of sea,” but the serpent’s body has been arched to form a third cryptogram: omega — the initial letter of Oceanus — to represent him as the grandfather of Thetis.21 Thus figure C is Tethys greeting Peleus, her future grandson-in-law, while her husband, Oceanus, introduces himself by curling into an omega.

There is, however, no Thetis in scene A; this leaves both the scene and the Catullan connection incomplete. Of the remaining figures, F — the showpiece figure of the entire frieze — seems the most likely candidate for the role, yet she is in scene B. Of course, the frieze may be one continuous narrative; then again, it may not.

Another complication arises for the viewer at this stage. Two mythologically unrelated tales of love form the central structure of Catullus 64: the happy story of Peleus and Thetis and the tragedy of Theseus and Ariadne. With scene A depicting the first of these, it is logical enough to expect scene B to depict the second or to interpret the frieze as one continuous scene depicting the first. However, the iota cryptogram, which is likely to function in the same manner as its pendant, would require scene B to be a location whose name begins with that letter, and there are no place names in either tale of 64 to fulfill that obligation. Thus, for some as yet unknown reason, the Catullan connection appears to terminate with scene A. Nevertheless, the apparent deviation is hardly a surprise. It is more logical to expect scene B to be as recondite as scene A. Merely to continue the Peleus and Thetis story or to go on to depict the Theseus and Ariadne story would pose no further challenge to the viewer.

Yet some valuable clues about scene B have been provided for the viewer. For example, although it would be natural to assume that the architectural cryptogram which begins the scene functions in the same manner as its pendant — supplying the first letter in the place name of scene B and the first letter of figure E’s name — the iota pier does not just serve as the initial “iota.”  It is also the second letter of the word for “pi” and, as I will show, is very much a part of that pi. One can see that unlike the pilaster in scene A, the iota pier is matched in color and perspective to the left-hand leg of the pi structure. It also displays clamp holes for attachment to an entablature, leading me to believe that it is to be interpreted as a disassembled member of that pi whose support function is assumed by Peleus. Since figure E obscures a portion of the iota in the same manner that figure A obscures the pi, then perhaps figure E can be construed as the second vertical stroke of the letter pi. In that case, the first letter of figure E’s name must also be pi. Futhermore, not only do figure A and figure E both hold the first position in each scene, but they are also very similar looking young men who are associated, as I have just shown, in both visual and abstract terms by means of the two architectural cryptograms. All of these correspondences lead me to conclude that figure E, like figure A, is also a mythological suitor whose name begins with pi.

As for the place — it has already been established that scene A is a mountain; since both scenes A and B show rocky piles and a three-limbed tree, then we can safely assume that scene B is also a mountain. So now that we have two suitors on two mountains we can further surmise that scene B, like scene A, also depicts a betrothal. And though I have just shown how we have deduced that the second suitor’s name begins with pi, the fact that his architectural feature is disassembled still allows us to conclude that the second mountain’s name begins with iota. If we consider the possibility that scene B may not be a radical departure from Catullus 64 but rather an interpolation of it, we might conjecture that scene B depicts a romantic tragedy similar to that of Theseus and Ariadne, an expectation that would likely have occurred to the Roman savant intimate with poem 64.

The betrothal of Peleus and Thetis on Mount Pelion is one of the two most famous in mythology; the other, directly linked to the first in a cause-and-effect relationship, is that of Paris and Helen on Mount Ida.22 Based on the preceding arguments and others to follow, we suggest that the designer of the frieze replaced the romantic tragedy in Catullus 64 with that of Paris and Helen.

Scene B, therefore, we identify as Mount Ida and figure E as Paris, seated atop a rocky pile just as he is often depicted in ancient art.23 As in scene A, this is the moment when the match is arranged.  In return for the Golden Apple, Helen is being promised to Paris by Aphrodite, figure G, whose eyes are firmly set upon him.

However, a definite quandary now exists. If we identify figure F as Helen, then there is no Thetis for Peleus and if she is Thetis, then there is no Helen for Paris; one bride and two suitors seems an impossible situation. But figure F cannot truly be Helen if she is only just being promised to Paris, for Helen is still in Lacedaemon with Menelaus.  Nor can figure F truly be Thetis because the scene is Mount Ida, not Mount Pelion. Furthermore, Peleus is in scene A, which, relative to scene B, lies in past time.  It is spatially and temporally impossible to identify the figure as either Thetis or Helen. Yet such a conclusion leaves figure F unidentified and the action of both scenes incomplete, for now neither suitor has his bride-to-be. The deadlock, however, may be resolved by reading the visual clues provided by the artist. Although positioned in scene B, figure F is depicted as both mentally and physically remote from it. Mentally, she seems completely unaware of the other two figures there; in fact, she seems to be asleep. It is particularly notable that F is the only figure in the entire frieze who is not looking at another. Physically as well, she is distanced from the other figures in scene B, positioned noticeably below and in front of them. Nevertheless, certain details in the frieze seem to associate figure F with each of the two suitors. Paris is definitely looking at the figure and the rocky outcrops upon which both are placed are shown as overlapping; Aphrodite’s, however, is set apart. Paris is positioned back from the edge of the frieze but Peleus and figure F are both positioned at the edge; furthermore, the torch-bearing Eros is leading him in her direction. The homology of the torch of Eros and the torch of figure F cannot be overlooked; the scarce attributes and the consistency of correspondence between scenes A and B virtually compel one to associate the two torches. The established import of the architecture finalizes the relationship of figure F with Peleus and Paris, for they are the only three figures associated with it: Peleus and Paris by their respective pilaster and pier, and figure F by the column base at her feet.

At this point there is ample evidence of both the exacting correspondence in these pendant scenes and of visual tropes to allow identification of figure F. With the exception of Eros, the scenes are identically composed. Both take place on mountains; figures A and E are suitors; figures D and G are Olympians. A correspondence, therefore, between figures C and F seems very likely. Thus the same duality exhibited by Oceanus in figure C is repeated in figure F by the visual rendering of a trope known as syllepsis. In literature, syllepsis occurs when one word or expression is made to perform two grammatical functions but only logically functions for one of them, as, for example, the use of are in Both she and we are willing. Although figure F is situated in scene B, which she logically completes as Helen, she also completes scene A as Thetis. Thus she performs a dual role, a dream figure who symbolizes both yet-to-be-realized brides: Thetis for Peleus and Helen for Paris. Her mental and physical detachment from her own scene, the languid pose, her state of reverie, and the wedding-torch slipping from her fingers all indicate this ethereal role.

Yet Paris rests his eyes upon the dream figure and surely sees only Helen; Peleus is on the other side of the vase and cannot see the figure at all. Again, correspondence assures the identification. Note how in both scenes the powers of love have their eyes firmly fixed upon the suitor. Paris does indeed see Helen before him, for her image is being conjured in his mind by the power of Aphrodite and transmitted by her unwavering gaze.24 Likewise, a vision of Thetis exists in the mind of Peleus, there created by the equally powerful eyes of Eros.

Tokens relevant to both Thetis and Helen are beneath the dream/bride figure. The column base that lies atilt at her feet and the flames of the nearby torch may symbolize the fire and ruin which the son of Thetis, Achilles, will bring to the Greek fleet and to Troy. These same tokens may also presage the fiery destruction which Helen will bring to Troy. The wedding-torch slipping from the bride figure’s hand is probably meant to be construed independently as well: a sign of tacit consent by Thetis that Peleus may approach; for Helen, that she will loose her marriage vows to Menelaus.

The frieze, therefore, is actually neither one scene nor two. Fully appropriate to the theme of the frieze, it is two engaged scenes, married by the bride figure, the pivotal character in both tales of betrothal depicted on the vase, tales that are themselves inextricably intertwined. Finer points in the composition of the frieze further integrate the two scenes. Each side of the vase is flanked on the left by a suitor; on the right, by an Olympian. The suitors face right; the Olympians, left. The framing structure of each scene, at first glance, seems identical. However, the bodies of figures E and G, unlike their pendants, are facing out from scene B, not into it, thereby creating a sense of openness and a visual cue prompting the integration of the two.

Six further details in the frieze induce the eye to see the frieze as an integrated whole: (1) The only figures positioned at the edge of the frieze are Peleus, Tethys/Oceanus, and the bride figure, who, as Thetis, ties the two scenes together via this link with her suitor and family. (2) The left foot of Paris extends beyond the iota pier into scene A, almost grazing the second tree there. (3) The truncated, finger-shaped limb on that same tree points directly from one scene into the other. (4) The entire ground line of the frieze is broken at one point only. As noted earlier, the iota pier is set well in the background. In marked contrast, the pi architecture which begins and ends the frieze is as far foreground as possible. The faithful correspondence of the two scenes indicates that if they were not to be treated as an integrated whole, the iota pier would also have been positioned at the edge of the frieze. (5) Both the acrostic nature of the word pi and the matched three-quarters view piers bond the two scenes. (6) The scowling, horned head between scenes A and B makes for a discouraging terminus. Just how critical it is to link the two scenes by means of the bride figure will be evident when the function of the three clamp holes is further discussed below.

The complex role of the bride figure, the architectural and figural tropes, the identifications of the two scenes and their figures, and recognition of the visual, causal, compositional, and chronological correspondences between the two sides of the vase are the prerequisite, essential clues provided by the artist in order to read the frieze at its primary level. Should an interpretation of the frieze be terminated at this stage, one is faced with several inconsistencies, none of which seems likely given the conceptually elaborate nature of the work: (1) The combined cryptograms have formed a nonsense word: pi omega iota. (2) A disparity in the action of the two scenes has been created for the viewer by the bride figure — visually, there are still two suitors and only one potential bride. (3) A temporal hiatus exists between scenes A and B — that is, relative to A, scene B, which includes the bride-to-be of Peleus, is in the future. (4) Ill omens lie at the feet of the bride figure. Simply stated, the two stories which constitute the narrative are incomplete.

As with the Spada Reliefs and certain pendant wall paintings at Pompeii, a “narrative of the interior, a product of the ruminations of a sophisticated observer who sees and moves on from one panel to the next” is now required in order to resolve and complete the two tales depicted in the frieze.25 Drawing on their intellectual corpus and understanding of the visual metaphors before their eyes, Roman literati would actualize visiones, internalized, imaginary visualizations, to complete in their own minds the work before them.26 The modern viewer can parallel and perhaps facilitate this process of internal narrative by visualizing the frieze as the first frame in a film which has been momentarily frozen in glass. But before the final pieces of this “erudite puzzle” are fitted, let us further study the frieze for clues that may help guide us through the process.

Scene A: From the position of his drapery, it is clear that Peleus was originally in front of the left-hand pier in the pi architecture and has just stepped right.27 Eros is leading Peleus to Thetis. Tethys greets and urges Peleus on with her right arm as Oceanus identifies himself by curling into an omega. A pensive Zeus ponders the surrender of a woman he himself desired.

Scene B: The bodies of Paris and Aphrodite are shown in tensed positions that indicate they are about to rise. Both figures have one of their legs drawn back in order to shift body weight to the ball of the foot. Aphrodite, her right arm stiffened and straightened, presses down upon the rock with her hand to facilitate the movement. Paris is already gathering up his drapery with his left hand.28 The way their bodies face indicates in which direction they will rise.

The architecture: To review, the matching three-quarters-view iota pier and the three-quarters-view pier of the pi structure are pure white. The thinly carved linear pilaster, so deeply shaded by the blue-black base glass, is thus imbued with an air of insubstantiality. The front and right sides of the iota pier’s capital display clamp holes; its lower half is obscured by Paris, just as the pilaster is obscured by Peleus. A detached column base with a central clamp hole lies at the feet of figure F.

To begin the internal completion of the narrative, we return, of course, to figure A. With Eros guiding him to Thetis, Peleus would now step away from the pi architecture. Because the pilaster is so dark and thin, the stark-white three-quarters-view pier and entablature create a fourth cryptogram — the letter gamma — the moment Peleus steps away. The first envisioned movement of the lead figure in this second level of interpretation has changed the theme of the frieze, in logical sequence, from betrothal to marriage, for the gamma must surely stand for gamos, the wedding. Architectural tropes again come into play, for we also observe that the decorative elements of the architecture continue the wedding theme by the union of male Doric and female Ionic elements on the entablature.29 Noteworthy, too, is the realization that the Doric triglyphs (a male element) are vertical, the Ionic dentil course (a female element) is horizontal, and that a correspondence with the suitors, who are upright, and the bride figure, who is prone, is perfect.

Seeking his bride-to-be, Peleus would now be approaching scene B, which, as it stands, is impossible for him to enter. When Peleus enters scene B, it cannot be Mount Ida, figure F cannot be an ambiguous bride figure, and the time plane cannot be the future. Scene B must be transformed. Paris and Aphrodite would now turn from each other and rise. Simultaneously, the vision of Helen in Paris’ mind dissolves. Aphrodite then vanishes right into scene A, there to await the upcoming nuptials of Peleus and Thetis with the other deities. When Paris rises left, he obscures the entire iota pier with his body.

The same mutable qualities we saw in the pi architecture would now occur in the iota pier. Its detached base returns and is affixed to it by the clamp hole in its center.

The entire structure, including Paris — for he now is the pillar, as Peleus was earlier, — would be repositioned as the right-hand portion of the architecture in scene A which has just been vacated by Peleus. The clamp holes in the capital lock the pier into the front and side of the entablature and the two matching piers, both three-quarters-view and squared, are now side by side. Because the image of Helen created by Aphrodite no longer exists, figure F is now Thetis. The entire frieze has become one place, Mount Pelion; one time plane, the present; and one continuous narrative. Peleus would now enter the other side of the vase. There, alone on her couch, waits Thetis. Eros, his task completed, flies on and joins his mother. The lovers tryst. Peleus then continues on and resumes his original position in front of the left-hand pier of the reunified piece of architecture.

It seems best to remind ourselves that only we, the viewers, are aware of the mutations of the architecture and of the changes wrought by our internal completion of the narrative. Because they are on opposite sides of the vase and on different time planes, the figures in scene A could never have known of the original scene B, or vice versa. In scene A, the linear pilaster in the pi architecture, barely perceptible to the viewer, could not have been perceived by the figures because it is only two-dimensional. In scene B, the figures have their backs to the iota pier, which is set well behind them. Nor could its detached base have been seen by them, for their view was obscured by a projecting ledge of the bride figure’s couch.

Thus our final image of the frieze shows Thetis on her wedding couch on one side of the vase and Peleus and the immortals awaiting her arrival on the other. Being on a different time plane, the figures are unaware that Paris is in front of the right-hand pier of the pi architecture. We the viewers, however, envision him and remark how the two tales of the frieze are now ingeniously amalgamated. We see a tableau vivant depicting the wedding of Peleus and Thetis which simultaneously presages its own consequences.

The artist has taken care to insure that the final image supplies its own proofs of a successful resolution of the scenes. Let us consider the remaining architectural cryptogram, the pi. To begin, its appearance is pleasing and balanced because it now consists of two piers matched in both color and perspective. Next, when Peleus served as the original right-hand columen of the pi, it was noted that his name begins with the letter pi and that he could be construed as forming one of the vertical strokes of the letter. In our final image, Peleus is before the left-hand pier of the pi and Paris is before the right. Now both vertical strokes of the pi structure are formed by the suitors, both of whose names begin with the letter pi. The architecture, as a result, has become a pi built from two pi‘s. Thus in abstract as well as visual terms, the cryptogram and its two piers are in perfect harmony. The final image also eliminates the nonsense word pi omega iota. In an impressive finale, the remaining two cryptograms create a new word, pi omega, and terminate the cipher precisely as it began, with a homonymic trope: pi omega is again an adverb in the Doric dialect meaning “Where?” and thus both repeats and plays upon the original trope. But it is also an imperative form meaning “Drink!” — a reward, no doubt, for one who solves the puzzle, for the Portland Vase “may well have been filled with wine on occasion.”30 Moreover, it seems entirely apropos that the terminus ad quem of the four cryptograms is the omega, that is, Oceanus. As line 30 of Catullus 64 tells us, the ancients believed Oceanus to be a great river of sea which encircles the world and returns into itself, just as the architectural cryptograms and their resultant homonymic tropes do; just as the narrative and our two circuits of the frieze do.

The concinnity of the artistic tropes that encode the frieze are subtly reflected in its overall theme. Be it betrothal on the first level of interpretation or marriage on the second, the theme of union, first hinted at by Eros, pervades the frieze from beginning to end, most poignantly illustrated by its operative key, the letter pi. The frieze begins with the architectural cryptogram pi, but, as we come to realize, the right vertical stroke — mate and match of the left — was sundered and used to form the letter iota, thereby producing a framework of visual tropes by which the frieze may be read. It is only when the two individual letters are united, and both the architecture and the letter pi are one, that the frieze itself becomes a unified whole. The thematic statement visually actualized by Eros (who dominates the frieze both by his placement above the other figures and by his immediate recognizability), that two become one, is exhibited six times by (1) Peleus and Thetis; (2) Helen and Paris; (3) the two three-quarters-view white piers; (4) the male Doric and female Ionic orders; (5) the two Doric adverbs, bonded by the superimposition of their initial letter as well as by meaning; and (6) the marriage of scenes A and B by means of the bride figure. The dual figures, Oceanus (river/grandfather) and the bride figure (Thetis/Helen) further complement the theme.

There is strong evidence that scene A depicts lines 26-30 of Catullus 64, but the issue of how both scene B and this interpretation as a whole relate to Catullus 64 has only been partially addressed. I suggest that not only scene A but also the structure, the mechanics, and the themes of the frieze throughout both levels of interpretation are drawn directly from Catullus 64.

There is a striking congruity in the arrangement of both poem and frieze. The basic structure of Catullus 64 is A-B-A: the beginning “A” section tells of Peleus’ and Thetis’ betrothal; the closing “A” section tells of their marriage. The “B” section is an inner story, the tragedy of Theseus and Ariadne.31 The architectural cryptograms, which function as both the physical and abstract dividing points in my tripartite interpretation of the frieze, delineate both the same wrap-around structure and sequence of events as the poem: scene A — pi: the betrothal of Peleus and Thetis; scene B — iota: the tale of Paris and Helen; scene A — gamma then pi: the marriage of Peleus and Thetis.

Another structural parallel of the two works invites comparison. Quinn notes that in the “A” section of Catullus 64 “the movement of the narrative is strictly linear, but the inner story [B] moves both backwards and forwards in time.”32 My interpretation of the frieze exhibits the same pattern. Scene A unfolds along one time plane only, the present, throughout both levels of interpretation. Scene B, however, begins with Paris looking forward in time at his vision of Helen. When Paris and Aphrodite rise, Paris momentarily returns to the present and then moves backward in time when relegated to scene A.

The dual roles of Oceanus and of the bride figure in the frieze also have antecedents in the poem; that of Oceanus has already been discussed. The bride figure, on the first level of interpretation, was simultaneously Thetis and Helen. The two brides, embodied in one figure, figuratively shared the wedding couch, center stage, in scene B. At the close of the internal resolution of the narrative, there was only Thetis. An identical schema occurs in the poem. As the opening “A” section of 64 comes to a close, Catullus focuses the reader’s attention on Thetis’ wedding couch in the center of a reception room. It is a key point in the poem; Quinn notes the “emotional energy built up by the tensioned description of the marriage-bed.”33 Upon the bridal couch is a coverlet embroidered with scenes from the tragedy of Theseus and Ariadne. A description of a scene on the coverlet featuring Ariadne then serves as the bridge to the “B” section. Thetis is the only bride in the closing “A” section. Thus the bridal couch of Thetis, figuratively shared by Ariadne on the coverlet, marries the two tales in the poem, just as the bridal couch in the frieze, shared by Thetis and Helen, marries scenes A and B. Both poem and frieze end with Thetis as the only bride.

I have already shown how the chronological, causal, and thematic links between the tales depicted in scenes A and B of the frieze are clear. In Catullus 64, however, no such ties are apparent between the “A” and “B” sections of the poem: the temporal relationship of the two sections is confounded, the tales are mythologically unrelated, and no overall theme to the poem is readily discernible.34 Whatever correlation Catullus may have intended between his “A” and “B” sections is unclear, for “there is an undeniable awkwardness about the structure of the poem.”35 Perhaps this “awkwardness” was apparent to the designer of the frieze, who then, for specific purposes, emended the poem.

The section of Catullus 64 containing lines 26-30 which is mirrored in scene A of the frieze takes the form of a double apostrophe.36 There is only one other double apostrophe in 64, in the “B” section, in which Catullus addresses Cupid (Eros) and Venus (Aphrodite).37 He states that by their powers, either rapture or misery can be mixed in the same cup. Perhaps this “love can be heaven or love can be hell” theme often found in Catullus’ other works was the one intended to bind the “A” and “B” sections of poem 64.38 If so, and if the designer of the frieze wished to remain true to that theme, then substituting the tragedy of Helen and Paris for that of Theseus and Ariadne would not alter the integrity of the poem but would produce some desirable results. Certainly a more logical and cohesive narrative is created by the temporal, mythological, and thematic relationships that exist between the Peleus-Thetis and Paris-Helen tales. Simultaneously, a work is created that would exemplify a favorite Augustan theme: the linking of the classical past to the Roman present. The Theseus-Ariadne tale is unconcerned with Troy or Rome; the Paris-Helen story begins a chain of events that leads to the fall of Troy, the rise of Rome, and, eventually, to the age of Augustus. By adding Eros in scene A of the frieze to Catullus’ cast in lines 26-30, and by replacing the “B” section of 64 with a tale that has Aphrodite as the driving force in scene B, the frieze now incorporates both double apostrophes of the poem and voices the theme of the second. Moreover, the divine ancestors of the Julian House are assigned key roles in each of the two scenes: Cupid leads Peleus to Thetis; Venus awards Helen to Paris. In light of the above, J. G. F. Hind’s speculation that the frieze may be “an early imperial essay in adapting Hellenic legend to relate to Rome’s past, and specially to Rome’s Augustan present,”39 is accurate as far as it goes. However, I think that it is especially to Rome’s future that the frieze relates and that its ultimate significance, as I will show, is encoded within a third level of meaning that personally and profoundly reflects the dynastic hopes of the imperial household in 25 BCE. As is the case with many Augustan works of art — the Ara Pacis for example — the Portland Vase is both commemorative and specific. To conclude that the vase, a private piece, was merely an “erudite puzzle” echoing a generalized public theme is absurd.

Harden writes that B. Ashmole’s “characterisation of the ancient status of the vessel would be accepted by everyone who has studied the piece: ‘The vase must always have been a precious object, made to order, and perhaps for some special purpose or occasion.’”40 Of course, and the occasion was unquestionably a wedding. Although Painter and Whitehouse, among others, are accidentally correct in their assumption that “the Portland Vase can only have belonged to the emperor Augustus himself,” the question remains: To commemorate which imperial marriage of the early principate was the vase made?41 By mending the flaws that exist in this interpretation as it now stands, the answer to that question will reveal itself.

Three flaws should be evident: (1) The pi cryptogram in scene A identifies both a person and a place, yet the iota cryptogram in scene B identifies only a place; Paris is identified by the relationship of the iota to the pi in scene A, not by the iota alone. The two architectural cryptograms, therefore, are inconsistent with one another. The precise correspondence of the two scenes demands that the iota also identify a person. (2) The detached column base, the foot of the pier in scene B, lies at the foot of the bride figure; this can hardly be accidental. Unlike the other two pieces of architecture to which it is integral, the role of the column base has been practically insignificant. The placement of the column base is far too visually prominent to play such a minor role in understanding the bride figure. (3) No explanation of the relationship between the two architectural units in scene B has been given.

As from the beginning, the encoded frieze reveals its secrets when the viewer heeds the flawless logic of the architectural cipher. Therefore, just as the detached right-hand vertical of the pi in scene A serves to identify a person (Paris), the detached column base must also serve to identify a person; in this case, one whose name begins with iota. The column base will then have executed the two functions required of the cryptogram while simultaneously maintaining the precedent established by the pi regarding a detached architectural unit. It is a straightforward and completely logical progression: the column base is part of the iota which is part of the pi. The lead figures of each scene (A and E) and the focal figure of the frieze (F) are the only figures to whom the architecture is visually and textually germane. That two of the three pieces of architecture would identify a person by the initial letter of their name and that the third would not defies explanation. Furthermore, because figure C is in fact a triple figure composed of a dual (Oceanus/river and Oceanus/grandfather) and a single figure (Tethys), figure F, by correspondence, must likewise be a triple figure.

The dual figure F is bride/Thetis and bride/Helen; the single figure F — who lies on Mount Ida in the presence of Aphrodite, divine progenitrix of the Julian clan by union with Anchises on the same Mount — I further identify as Julia (Latin Iulia), daughter of Augustus and bride of Marcellus in 25 BCE.

However, if there are three brides, then there must be three suitors. As described above, the final image would show the two suitors, each before a vertical stroke, or pila, of the pi architecture in scene A. Assuming that the viewer, at the very least, would spend several hours unravelling the complexities of the frieze, the figures become virtually ‘burned’ into one’s consciousness and are therefore internally visualized in the same positions as in the frieze (see illustration). Note how the lead leg of each suitor meets that of the other, and that the legs of each are in precisely the same position, from the knee down, but with left and right reversed so that their lead legs can indeed meet. I emphasize from the knee down because that is the point at which Paris’ incorporation into the architecture, and thus his role within it, begins. The fact that the lead leg of Paris now meets that of Peleus further explains why, in the original scene B, Paris’ leg extends beyond the iota column into scene A. Thus joined, the legs of the two suitors, from the knee down, create the final cryptogram — mu — for Marcellus, the third suitor.

Just as Julia is incorporated into the other two brides, so Marcellus is incorporated into the other two suitors. Marcellus, the man upon whom the future of Rome was intended to stand, becomes a composite of the two mythological suitors: Peleus, the most worthy mortal, and Paris, the man who married the fairest of the fair; who was esteemed by the gods for his prudence and sagacity; who awarded the golden apple to Julia’s divine ancestor; who was received by Priam, King of Troy, as his son and who, by his marriage to Helen, set the stage for the founding of Rome, and the future empire to which Marcellus was to be heir.42

Note that the mu cryptogram of Marcellus is superimposed upon the iota cryptogram of Julia. Because they are the only couple actually joined in the frieze (Peleus’ Thetis is still on the other side of the vase; Paris’ marriage is in the future), Julia and Marcellus are abstractly spotlighted by the final image. They hold and share the first position in the final image and do so within a structure displaying both male and female architectural details. If, in the final image, the architecture were to contain only the two male suitors, the absence of a female would create imbalance and disharmony. All the architectural building blocks that dominated and directed the action of the frieze have joined to create a stable, unified whole reflecting the dynastic hopes of Augustus for a stable and unified empire. And, not least important in a culture fond of allusion and rife with superstition, the total number of gamoi has been raised to a propitious seven, which (coincidentally?), matches the number of figures in the frieze.

Full resolution achieved, the final image draws a profound contrast between the human and the divine (see illustration): Bound to and by the small Doric shrine created by the architecture, the mortals are assembled within, segregated from the gods who stand apart. The gods, as Catullus states both midway and in the final two lines of poem 64, do not congregate with mortals.43


For their reading and criticism during various stages of this paper, I wish to acknowledge and thank Laetitia LaFollette, Elizabeth Will, Gilbert Lawall, David Grose, Richard Brilliant, Bettina Bergmann, Vincent Cleary, Audrey Houghton, David Whitehouse and Brian Cook. However, the views expressed in this article are mine solely and do not necessarily reflect the views of the readers.

1.         The entire Journal of Glass Studies 32 (Corning, N.Y., 1990), [JGS], is devoted to the Portland Vase; Appendix V is a schema of previous interpretations.

2.         Haynes’ and Simon’s interpretations are compared and analyzed in JGS, p. 130ff.

3. Glass of the Caesars (Milan, 1987), pp. 62-63.

4.         (a) Most scholars date the Portland Vase 25 b.c.e.-25 c.e. Painter and Whitehouse are more specific, 30-20 b.c.e. “or very soon afterward,” JGS, p. 123. (b) The Power of Images in the Age of Augustus (Ann Arbor, 1988), p. 253.

5.         (a) “The rarest and most elaborate luxury vessels of the early Roman Empire are cameo glasses,” D. Whitehouse, Glass of the Roman Empire (Corning, N.Y., 1988), p. 24. (b) “Carving a cameo glass required . . . a considerable investment of time”; JGS, p. 126.

6.         P. 253.

7.         For example, N. Horsfall writes that Vergil enjoys puzzling his learned readers with such “games” as “Unscramble the anonymous mythological reference” in “Vergil, Parthenius and the Art of Mythological Reference,” Vergilius 37 (1991).

8. Visual Narratives (Ithaca and London, 1984), p. 53.

9.         (a) The abstruseness of Catullus 64 has prompted such comments as “Catullus 64 is, in many ways, a bewildering poem, a maze difficult to penetrate without a plan of some sort”; The Cambridge History of Classical Literature, vol. II, Latin Literature (Cambridge, 1982), [CHCL], p. 188. (b) For Catullus’ sparing use of epithets and propensity to unusual ones, see N. Moseley, Characters and Epithets (New Haven, 1926), pp. 6-7.

10.        References to Catullus 64 are nothing new in respect to the Portland Vase frieze. No other interpretation, however, derives directly from the poem, as their identifications so prove. For example, Haynes states that “a romantic version of the myth [of Peleus and Thetis] close to that which inspired Catullus’ poem [64] must lie behind the frieze of the Portland Vase”; The Portland Vase (London, 1964), p. 19.

11.        According to E. Simon, Die Portlandvase (Mainz, 1957), scene B, the initial rendezvous of Apollo and Atia, Augustus’ mother, is the first of the two. This seems both visually and chronologically incompatible with the natural flow of the frieze.

12. De Architectura, 1.ii.6.

13.        Ibid., IV.iii.4.

14.        Ibid., VII.v.5-7.

15.        See Suetonius, Divus Iulius, LVI.6; Divus Augustus, LXXXVIII; and Aulus Gellius, XVII.IX.1-6.

16.        The well-read Roman was familiar with acronyms and acrostics, the first extant example of which is in Aratus, Phaenomena, 783-87, spelling lepte, “refined style.” Germanicus, Avienius, and Cicero (who praises Aratus in De Oratore, I.xv.69), all translated the poem into Latin. See also G. Hersey, The Lost Meaning of Classical Architecture (Cambridge, MA: MIT Press), 1988, p. 4: “Trope dwells in the world of puns, homonyms, and associations. It is playful and poetic, not scientific, and it is often etymologically incorrect. But it is unquestionably the way the ancients, including Vitruvius and his contemporaries, thought about words. (My emphasis.)

17.        Op. cit., IV.i.6-7.

18. Thessaliae columen Peleu, cui Iuppiter ipse,

ipse suos diuum genitor concessit amores;

tene Thetis tenuit pulcerrima Nereine?

tene suam Tethys concessit ducere neptem,

Oceanusque, mari totum qui amplectitur orbem?

19.        See Appendix V in JGS.

20.        K. Quinn, Catullus: The Poems (London, 1973), p. 306.

21.        In the chapter “Pendants and the Mind’s Eye,” Brilliant quotes Quintilian and writes, “It would appear that the extraction of elements from their cohesive narrative contexts and their subsequent employment to represent themes of greater or lesser degree of abstraction is an exercise in modal forms, expressed in linguistic or visual terms.” After discussing visual application of the metaphors, synecdoche, and metonymy in wall paintings in the Theban Room, House of the Vettii, he writes, “The ambivalent possibilities of reading [the pictorial narrative] can hardly have been coincidental”; p. 73.

22.        The appearance of the Golden Apple at the wedding feast of Peleus and Thetis led to the Judgment of Paris and his subsequent marriage to Helen.

23.        (a) For an example see P. Mayerson, Classical Mythology in Literature, Art, and Music (New York, 1971), fig. 85, p. 390. (b) A figure unquestionably identified as Paris is carved upon a cameo disc that was set into the base of the Portland Vase in antiquity. The research of Painter and Whitehouse has shown that the Paris on the disc-base “was aligned directly under figure E, hinting at the identity of the two”; JGS, p. 135.

24.        “I saw your features in my mind before I saw them with my eyes,” Paris writes to Helen in Ovid’s Heroides, XVI.37 (ante tuos animo vidi quam lumine vultus).

25.        Brilliant, p. 89. See also See P. H. Blanckenhagen, “Der Erganzende Betrachter, Bemerkungen zu einem Aspekt hellenistischer Kunst,” in Wandlungen: Studien zur antiken und neuerer Kunst, Ernst Homann-Wedeking gewidmet (Waldsassen-Bayern, 1975), pp. 193-201.

26.        Quoting Quintilian, “There are certain experiences which the Greeks call phantasia and the Romans visiones whereby things absent are presented to our imagination with such extreme vividness that they seem actually to be before our very eyes,” Brilliant discusses this process in detail, pp. 66-67, 77-78.

27.        Cp. Haynes: “the way in which the lower folds of the cloak in his right hand cling round the foot of the nearer pillar proves that he is picking the garment up, not throwing it off”; p. 14.

28.        Cp. Haynes: “The fingers of his left hand toy with the edge of his cloak”; p. 15.

29.        Vitruvius describes the ancients’ gender associations with the architectural orders, IV.i.6-7. See also G. Hersey, passim.

30.        Haynes, p. 20.

31.        See Quinn for a complete structural outline of the poem, pp. 298-99.

32.        Quinn, p. 299.

33.        See his note regarding lines 50-51, “Bridge Passage,” p. 309.

34.        For the temporal problems in the poem, see Quinn’s note regarding line 11, p. 302.

35. CHCL, p. 188.

36.        Lines 22-30.

37.        Lines 94-98.

38.        So Quinn suggests, p. 309.

39. Glass of the Caesars, p. 63.

40. Glass of the Caesars, p. 64.

41. Roman Glass: Two Centuries of Art and Invention, edited by Martine Newby and Kenneth Painter, vol. XIII of Occasional Papers from the Society of Antiqueries of London, 1991, p. 41.

42.        For those who may feel that Paris is an inappropriate paradigm, see R. Brilliant, Visual Narratives, p. 59.

43.        Lines 267-68 and 407-08.

1. Scene A of the Portland Vase frieze. Reproduced courtesy of the Trustees of the British Museum.
1. Scene A of the Portland Vase frieze. Reproduced courtesy of the Trustees of the British Museum.
2. Scene B of the Portland Vase frieze. Reproduced courtesy of the Trustees of the British Museum.
2. Scene B of the Portland Vase frieze. Reproduced courtesy of the Trustees of the British Museum.
3. Figure A of the Portland Vase frieze. Reproduced courtesy of the Trustees of the British Museum.
3. Figure A of the Portland Vase frieze. Reproduced courtesy of the Trustees of the British Museum.
4. Detail of the Portland Vase. Reproduced courtesy of the Trustees of the British Museum.
4. Detail of the Portland Vase. Reproduced courtesy of the Trustees of the British Museum.
5. Detail of the Portland Vase frieze. Reproduced courtesy of the Trustees of the British Museum.
5. Detail of the Portland Vase frieze. Reproduced courtesy of the Trustees of the British Museum.
6. Figure A and Figure E. (Author's revised rendering.)
6. Figure A and Figure E. (Author's revised rendering.)
Reproduced courtesey of the Trustees of the British Museum.
Reproduced courtesey of the Trustees of the British Museum.
Author' reviised rendeering
Author' reviised rendeering