Perhaps it is fair to say that we readers of the Commedia do not have a wide enough perception of our poet's confrontations of practical problems, problems that have little to do with instruction or delight, doctrine or bellezza, but those stubborn, difficult tactical decisions necessitated by the very fiction created by its writer. For example, once this poet had decided upon the presence of "rivers" as borders, separating zones of the geography of hell, he had either to borrow the methods in which his protagonist crosses these barriers (e.g., he might choose to remember Aeneas on the Styx as precursor of our hero's transit of the Acheron) or he might choose to invent his own (e.g., his protagonist's crossing a latter-day version of the Styx on a Phlegyas-propelled skiff, with the detail, borrowed from the crossing of Acheron in Aeneid VI, that Dante’s weight makes this conveyance descend under the protagonist’s weight in imitation of the Virgilian scene, suppressed in Dante’s third canto only to be resurrected now in his eighth). On the other hand, once the poet had to deal with his own invented rivers, Phlegethon and Cocytus, he apparently felt that he had been left to his own devices. Thus Nessus is obviously a "poetically necessary" replacement for a watercraft, the poet having exhausted his supply of familiar Classical references in the "ferries" operated by Charon and Phlegyas in Cantos III and VIII. (He will have recourse to Antaeus as a fourth and final Infernal "ferry" down to the ice of Cocytus.)

Verses 126 and 139 of Inferno XII mark the confines of an extraordinary moment in the narrative, in which the protagonist first mounts Nessus, then rides him across the lowest part of Phlegethon, and then dismounts and watches the centaur re-cross the river.[1] A less accomplished artist – or at least one possessing less in the way of self-control – would have rendered these actions more effusively, would have borrowed more bounteously for this imposing scene, in which a Classical centaur, Nessus (Ovid, Metam. IX), enters this Christian historical narrative as easily as though he actually belonged in it. Perhaps nowhere else in this poem are actions described less animatedly: e quindi fu del fosso il nostro passo (126) and (of Nessus now) poi si rivolse e ripassossi 'l guazzo (139). It is difficult to know exactly what is happening, even who is carrying whom.[2]

The scene contained between these points is observed by the protagonist – as some readers do indeed realize – from his elevated temporary seat upon the back of Nessus. Nowhere, however, does the poet tell us directly that he has mounted (or dismounted) the centaur. He leaves it to us to make and to insist on such observation.[3]

In contrast to most participants in a commentary tradition on this passage that is centrally marked by its confusion, the brief gloss of Bosco/Reggio (comm. [1979], Inf. XII.94-96) is admirably clear and to the point: "Porti [...] groppa: solo da questa richiesta di Virgilio, sappiamo che Dante passa il guado del Flegetonte in groppa a Nesso, perché nel resto del racconto il particolare è taciuto." Amidst the surprisingly large number of commentators who simply seem not to understand the details of the physical action of the little scene, at least several modern commentators are aware that our poet is challenging us, his readers, to wonder what he is up to. Whatever the poet himself thought of this request and of its potential fulfillment in his text, he shows us neither a nervous Dante mounting (with Virgil's help? with Nessus going down on his bent forelegs to make the mounting possible?) nor dismounting at the beginning of the next canto. At verse 100 Dante and Virgil are apparently still moving on their own feet: "Or ci movemmo con la scorta fida." 

Hardly any of the first commentators come right out and say (Benvenuto da Imola and Francesco da Buti being two important exceptions) that the protagonist crosses Phlegethon on Nessus's back, if, within the givens of the poem, it is impossible that he did not (unless we are to understand that it was Virgil who took the protagonist across on his back), since, as a mortal human, he would have been destroyed in that river of boiling blood.

(1)Virgil makes a request for Chiron's aid in finding an agent of transport (XII.94-95):

e che ne mostri là dove si guada,
e che porti costui in su la groppa [...].  

(2)Virgil instructs the protagonist to prepare to mount Nessus (XII.113-114):

Allor mi volsi al poeta, e quei disse:
"Questi ti sia or primo, e io secondo."  

(3)The protagonist's new point of view (XII.115-116):

Poco più oltre il centauro s'affisse
sovr' una gente […]

(4)The episode ends with Nessus's departure (XII.139):

Poi si rivolse e ripassossi 'l guazzo.

It is probably impossible to argue that the reader is not authorized to understand that the protagonist is carried across Phlegethon by Nessus. Our only possible question is why the details are presented so undramatically, that more is not made of them. The only likely hypothesis in response to that question is that the poet did not want to risk too much of his credibility with such outrageous details, choosing instead to let us supply them for him. I find that I am echoing Carlo Grabher, who was, I think, the first commentator to understand the most important reasons behind the poet's decision to insist on this detail: "Della sua cavalcata sulla groppa del Centauro il Poeta non parla. È bastato il sobrio cenno di Virgilio: e che porti costui in su la groppa (95). Così l'artista ha dato uno spunto alla fantasia, ma ha poi evitato di svilupparlo nei particolari, che qui gli apparivano o banali o inutilmente digressivi." On the other hand, I do not agree with Grabher's conclusion: Dante knew full well what he was doing and is in fact inviting his reader to play along with him. The suppressed details are not banal, but closer to astounding: we are to understand that the protagonist has somehow managed to mount the back of this incredible beast. . . .

What I find extraordinary in the arguments of those who deny, by their silence on the issue, that the protagonist is meant to be understood as mounted in this scene, is that not one of them even mentions this hypothesis even to attack it, despite the fact that it has appeared in every period of the exegetical tradition (see the table immediately below). That looks to me like a form of discomfort with the argument, so troublesome that it may not be broached, even in disagreement.

The following is a table listing those commentaries on the twelfth canto found in the Dartmouth Dante Project who support the notion that the protagonist crosses the Phlegethon upon Nessus's back:

1. Benvenuto da Imola (1375), vv. 113-114.
2. Francesco da Buti (1385), vv. 76-139; 91-99; 127-139.
3. Anonimo fiorentino (1400), v. 139.
4. Guiniforto delli Bargigi (1440), vv. 91-99; 113-114.
5. Giovan Battista Gelli (1541), v. 139.
6. Alessandro Vellutello (1544), vv. 91-99.
7. Pompeo Venturi (1732), v. 139.
8. Baldassare Lombardi (1791), v. 139.
9. Luigi Portirelli (1804), v. 139.
10. Gabriele Rossetti (1826), vv. 85-99; 91-96.
11. Raffaello Andreoli (1856), v. 139.
12. Giacomo Poletto (1894), v. 139.
13. John S. Carroll (1904), vv. 83-100.
14. Carlo Steiner (1921), v. 139.
15. Carlo Grabher (1934), vv. 85-96, 139.
16. Luigi Pietrobono (1946), v. 139.
17. Giuseppe Giacalone (1968), v. 139.
18. Bosco/Reggio (1979), 55-100, 94-96, 101-135.
19. Robert Hollander (2000), vv. 114-115, 139.

I will permit my earlier self a final word: "In handing Dante over to Nessus, Virgil does not tell him to 'mount up.' Yet this is what we should almost certainly understand is happening. The rest of the canto, until its final line, takes place with Dante looking down from Nessus' back. This is almost clear when we consider the next verse: 'A little farther on the centaur stopped.' Up to now the movements of Dante and Virgil have been noted as they made their way along; now it is the centaur's movement which is recorded. Why? Because Dante is sitting on his back. Why did the poet handle this part of the journey so delicately? Perhaps because he was aware that the scene would have been 'outrageous,' a Dante on horseback in hell, too much for a reader to accept. In such a view, he once again invites the reader to become his accomplice in making his fiction."[4]

[2] Benvenuto (comm. [1375], Inf. XII.113-114) has difficulty dealing with the straightforward (if not anything like clearly described) details of the scene, and ends up with Virgil also mounted on Nessus for the crossing of Phlegethon.

[3] Hollander (DDP, comm. [2000], Inf. XII.139), notes a detail that is perhaps worthy of attention: "His task accomplished, Nessus crosses back over the river of blood without a word. It seems clear that he would not have crossed over [in the first place] had he not been carrying Dante on his back."

[4] Hollander (comm. [2000], Inf. XII.114-115).