




































Humanist Studies & the Digital Age, 3.1 (2013) 
ISSN: 2158-3846 (online) 
http://journals.oregondigital.org/hsda/ 
DOI: 10.5399/uo/HSDA.3.1.1596 
 

 

 

 

  53 

The Princeton Dante Project 

Robert Hollander, Princeton University 
 

History: The Princeton Dante Project (henceforth PDP) was the direct result 
of a suggestion by a former student of mine, Susan Saltrick (Princeton ’78), that 
I should meet with Bob Stein, at that time (late 1994) one of the most 
impressive innovators in the field of so-called humanities computing, with the aim 
of developing a second large-scale Dante project. My initial response to Susan 
was that I was too busy with the first, the Dartmouth Dante Project (DDP), for 
which I had been serving as Project Director since 1982, even to think about a 
second. (The DDP was a communal effort, as would be the PDP, involving 
people on both sides of the technological divide, professors of literature in the 
Department of French and Italian at Dartmouth and computing personnel from 
that institution's Kiewit Computation Center.) Susan convinced me that no harm 
could possibly come from spending some time with Mr. Stein. And so the three 
of us did meet at his office in Manhattan, along with some of his associates. It 
is sufficient to report that, from the expert advice and encouragement I received 
from Saltrick, Stein, and Stein’s collaborators, I was inspired to conceive a new 
project that would do many of the things for a student of Dante that the DDP 
was obviously not designed to do. That earlier project was tasked with doing a 
fairly simple chore: making machine-readable the most notable commentaries to 
the Commedia in the six hundred and sixty years separating the first 
commentaries from the current ones (1322-1982; we would eventually add three 
still more recent ones) and building a mechanism to search them effectively. 
Now that we had made significant progress with establishing that first Dante 
database, it was almost inevitable to envision an electronic edition of the poem 
with facing English translation (concerning which there will be more in the next 
paragraph). This “edition” would also have the following links: explanatory notes 
for nearly every passage in the poem; philological notes for verses containing 
problematic textual readings; and historical notes for things, places, and persons 
referred to in a particular verse. In addition, there would be links to all 
commentators in the DDP (then several dozen, now some seventy-five) who had 
dealt with the particular verse or passage under consideration. There would also 
be links to an Italian voicing of the poem, broken into tercets (the three-line 
unit that is the basis of the poem’s structure): the user would have the option 
of hearing a single tercet, or the screen “page,” or the entire canto. Finally, we 

http://journals.oregondigital.org/hsda/
http://creativecommons.org/licenses/by-nd/3.0/


Humanist Studies & the Digital Age  Hollander 

3.1 Spring 2013  54 

would offer a small sampling of illustrations of various scenes in the poem, 
including the most celebrated series of these, that of the great nineteenth-century 
illustrator Gustave Doré, but also the much less well known but striking 
lithographs, one from all but one of the hundred cantos, by the twentieth-century 
artist Amos Nattini. (The Dalì illustrations are also available by connecting to 
another site, as are a number of others.)  In addition, there is a collection of 
maps and diagrams; these are not keyed to specific passages but serve as a 
sort of appendix. There were to be (and are) other features as well: Paget 
Toynbee’s brief but full biography of Dante, along with a 1500-item bibliography 
of critical work referred to in the notes to the poem, as well as summaries of 
the contents of each canto. 

The question of which translation to use was resolved by my wife, the poet 
Jean Hollander. I was on leave in Florence for the spring semester of 1997. 
We were living on the grounds of the former Machiavelli estate, across the 
Arno, with an enviable prospect of the center of the city, at eye level with the 
dome of the Duomo. One day in February, soon after our arrival, my wife 
found me looking over a piece of John Sinclair’s translation, which I had chosen 
as the one we would use for the PDP and which we (I and graduate students 
Frank Ordiway and Lauren Scancarelli Seem) had modernized, getting rid of the 
worst cases of nineteenth-century poetic diction (e.g., "thee" and "thou," which 
both became "you"). When Jean examined a piece of that text she declared it 
to be unsayable. I reminded her that she had on occasion taught Dante to 
students using Sinclair’s translation and challenged her, asking if she thought she 
could do better. Her answer was to pick up the several pages of the first 
canto of the poem and carry it to her work space in that former Machiavelli 
dovecote in which we were so happily ensconced. A few days later she handed 
me a similarly-sized document, her revised version of Sinclair’s prose, now 
become free verse in Jean’s reshaping. What she presented was, for all its 
problems, a clear improvement over the original. I asked Jean if she wanted to 
translate the entire poem with me. She said she did, even after I warned her 
of how much work she was taking on. We published the first volume of our 
translation with Doubleday in 2000, acknowledging our debt to Sinclair several 
times in an opening note. In the successive versions after the first that we did 
of each canto, Sinclair disappeared from our view as we worked from Dante’s 
words rather than his translation. On the other hand, we still insist on our 
original debt to Sinclair, even as we did in that prefatory note eleven years 
ago. 

In the winter of 1996 I made application to the NYU “computer boot camp” 
directed by Red Burns, a faculty member at NYU, in association with Bob Stein. 
I was accepted and spent most of the month of June commuting by rail from 



Humanist Studies & the Digital Age  Hollander 

3.1 Spring 2013  55 

Princeton to Greenwich Village, where I had in fact grown up. The project went 
extremely well. The first day I drew on the blackboard a model of the design 
of the project as I foresaw it. I wish that I had made and kept a copy of that 
diagram, for that basic design was close to what we eventually produced -- a 
disk with representations of the various components of the PDP. I am grateful 
to everyone connected with that boot camp, teachers, trainees at NYU, and 
fellow students alike. That experience ensured that the PDP began its life on 
the best imaginable footing. When I went to talk to Princeton’s humanities 
computing people, then led by Jacqueline Brown, the reasons for creating the 
PDP practically explained themselves. To my great pleasure, Princeton wanted to 
undertake the project. And now we needed to solve the least interesting but not 
inconsequential part of the problem: money. 

The National Endowment for the Humanities (NEH) had funded, bless its 
heart, the DDP at a time in which few humanists (at least few serving on NEH 
panels) knew anything about computers (1983). NEH had earlier funded one part 
of David Packard’s huge Greek project, “Ibycus,” which eventually resulted in the 
computerized Thesaurus Linguae Grecae. The Endowment’s official position was, 
perhaps understandably so at that time, that humanities and computers were a 
marriage that needed to be studied more carefully before federal dollars should 
be advanced in their further support. It is surprising (as I know because I was 
allowed to read, as were all applicants in those days, the anonymous versions 
of the panelists’ responses  and found that these were as untutored as one 
might have feared) that the project came through with funding. I think (this is 
merely my conjecture) that the then-director of the NEH’s Division of Research, 
Harold Cannon, had seen the promise of the DDP and had shepherded the 
proposal through a very rocky pasture indeed. We later received further funding 
from NEH for a necessary second cycle during 1985-88. To my surprise (and 
no little delight) Dartmouth managed to put the whole thing together for an 
opening ceremony, with, as I recall, a dozen commentaries running on their 
mainframe, on October 8, 1988. . My most amusing memory of that celebrative 
weekend is the response of a visiting Dantean rival, a person who had already 
frequently not been impressed by my work, who peered into a computer screen 
and said, exasperatedly, “Oh, it's just Benvenuto da Imola,” as though it were 
normal to look at a screen and see a fourteenth-century commentary to Dante, 
as though the whole project were a step backward from going to the university 
library and opening up a volume of the five-volume set of the Vernon/Lacaita 
edition of 1887. The DDP, in the nearly quarter century since it opened to 
consultation from around the world, has in excess of a million contacts per year 
from, in a given month, more than fifty countries, led, unsurprisingly, by Italy. 



Humanist Studies & the Digital Age  Hollander 

3.1 Spring 2013  56 

That was then and the PDP was now. I thought the proposal I had drafted 
was the best one I had ever submitted for financial support. This time, however, 
the NEH response was not only negative, but firmly so. I thought that the PDP 
had been held hostage to its successful elder sister. Or perhaps my view of 
the younger sister overrated her charms. At any rate, Princeton’s financial 
involvement "paid" for more than Dartmouth's and the total amount of cash 
required was, most fortunately, not nearly so much. The DDP eventually had 
cost over a million dollars in cash from a number of foundations and individuals. 
The PDP was much less exigent with regard to outside funding for four 
reasons: (1) digitization of materials had become less expensive in the 
intervening dozen years; (2) there were far fewer materials to be made machine-
readable in the first place; (3) Princeton put more of its own resources into this 
effort than Dartmouth had, mainly a goodly piece of time from five or six of its 
employees in computing; and (4) most of the project was essentially completed 
within a single academic year. Major support for the PDP came from the 
Andrew W. Mellon Foundation (as is frequently the case, the first to see merit 
in a challenging proposal), the Gladys Krieble Delmas Foundation, and the 
Edward T. Cone Foundation. Within Princeton the project benefited from a grant 
from the 250th Anniversary Fund for Innovation in Undergraduate Education. In 
2000 there were two substantial gifts from individuals: George Castell and Paul 
Gridley. (Mr. Gridley also helped to fund the renovation of the Dartmouth Dante 
Project in 2004.)  The Clover Foundation provided an emergency grant to ensure 
the completion of the first stage of the project in the fall of 1998. Where the 
DDP had had to raise more than $1,000,000 in cash, the PDP needed roughly 
one fifth of that amount, since the University supplied, as its contribution, so 
much of the necessary brainwork at no charge to the project. 

There were nine of us charged with designing and building the PDP and we 
met fairly frequently during the course of that academic year. The cast of 
characters included four people who worked for CIT (Computing and Information 
Technology -- now OIT, Office of Information Technology): Kirk Alexander (School 
of Engineering), Peter Batke (CIT), James Beldock (independent software 
consultant), Paula Brett (CIT), James Chu (CIT), Robert Hollander (French and 
Italian), Catherine Kunicki (independent designer), Lauren Seem (graduate student 
in Comparative Literature), and Jonathan Wilding (CIT). There were several 
others, as well, including Kirk Alexander’s younger colleague, Kevin Perry. He 
and Paula Brett are the main contacts for the few items of housekeeping that 
are required now, a dozen years later, on an ongoing basis. Indeed, I spent a 
good part of the early summer of 2011 reviewing and updating with the two of 
them the entire database, some of which had fallen into disrepair or 
obsolescence. 



Humanist Studies & the Digital Age  Hollander 

3.1 Spring 2013  57 

Function: Since the PDP offers a tutorial that users may consult and since 
much of what the database offers is straightforwardly intuitive, those who have 
never or rarely visited the site are here invited to have a look at the online 
tutorial or, if they prefer, at the project itself. The tutorial runs fewer       
than thirteen minutes and will give a concrete sense of what is       
contained in the database and how one consults those data: 
http://etcweb.princeton.edu/dante/tutorial/index.html. 

Some observations about the project’s functionality follow. In 2004 I 
experienced an illness that left me housebound for several months. I would not 
have been able to “do Dante” in the way that I have learned to trust and 
enjoy without the PDP -- handsome remuneration for my share in the toil 
necessary to create it. Most of the work that I have published on Dante is 
based in the close study of a particular text or series of texts. (Given that 
predilection, it will not seem accidental that I have written a commentary on the 
Commedia -- it also is contained in the PDP.)  Thus the PDP reflects its 
designer’s way of conceptualizing his own way of approaching this great poem. 
It is important to know that Dante considered himself not only the maker of his 
poem, but also its first commentator. In addition to its wider concerns, The 
Epistle to Cangrande (1318?) is in part dedicated to analysis of the first sixteen 
verses of the Paradiso. It is true that, in the years between 1819 and the 
present, more than a few, those who do not think the lines of argument found 
in the epistle are “Dantean” in nature, have raised their voices in dissent. The 
reader should know that the present writer is of the strong opinion that the 
document was, in fact, composed by Dante (see my polemically titled Dante’s 
Epistle to Cangrande, Michigan 1993) and is in accord with what the poem 
exhibits as its own instruction on how it should be interpreted -- a view that is 
increasingly shared among Dantists, if not by the few remaining hard-line 
dissidents. Whether or not Dante was the author of the first one, no other 
literary work in the Western tradition has been the focus of so many line-by-line 
commentaries. And that fact tells us something about its author and his method 
of writing. Dante, in fact, practiced self-exegesis (he was perhaps the first in 
Western letters to do so) in two of his preceding works, the Vita nuova and 
the Convivio. It is thus not at all surprising to find him doing so, if 
fragmentarily, in response to his own Commedia. 

The focus of the PDP is on the meaning of the smaller units of the poem 
-- a single word, a single verse, a single tercet. This is not to suggest that 
larger issues or themes -- empire vs. republic, the role of the papacy in the 
political world, the efficacy of prayer, the freedom of the will, etc., etc. -- do 
not come frequently into play, both in the poem and in the notes that 
accompany it, but that larger concerns must always be approached with the 

http://etcweb.princeton.edu/dante/tutorial/index.html


Humanist Studies & the Digital Age  Hollander 

3.1 Spring 2013  58 

details contrived by the poet much in mind. The three kinds of notes available 
in the text of every canto (interpretive, historical, and philological) reflect different 
concerns but are all appropriate to nuanced study. (Currently only the first six 
cantos have philological notes, but the rest are coming in the next two years, 
as part of my next project, a three-volume study of 101 cruces in the poem, 
one from each canto.) The entire project also benefits from the presence of all 
of Dante’s less known and revered works, not only the Vita nuova, De vulgari 
eloquentia, Convivio, and Monarchia, but his nearly one hundred vernacular lyrics, 
his four Latin eclogues, his thirteen epistles, and a treatise (the Latin Questio 
de aqua et terra). All of these are accompanied by standard English translations. 
And all of them are searchable by the search mechanism that was designed 
into the project early on, mainly, if memory serves, by Peter Batke. 

The historical notes all derive from Paget Toynbee’s Concise Dictionary of 
Proper Names and Notable Matters in the Works of Dante (Oxford: Clarendon 
Press, 1914), as revised by Charles S. Singleton (Oxford: Clarendon Press, 
1968). These notes remain the single most helpful assemblage of such material 
in the PDP. The fact that, for many items, the user is able to choose between 
a shorter and a longer version of the material turned up in a search is a 
boon. Let us examine the situation confronting someone who wants to know 
more about the mysterious prophecy of the “five hundred ten and five” 
(cinquecento dieci e cinque; Purg. XXXIII.43). This example also serves to 
illustrate a potential problem facing users of this mechanism in the database. If 
one types in these four Italian words in the “Toynbee Title Search” box, one 
gets a null result. However, typing in the same four words in the “Toynbee Text 
Search” box takes the user to the internally coded entry “DXV” (the Roman 
numerals for the number 515 by which scholars refer to this particular bit of 
business. The PDP does not do all the users’ thinking for them and there are 
and always will be some rough edges). Generally, once one has a “hit,” the 
default value for display is the condensed version, the “summarized entry,” brief 
and to the point; it may give searchers who want to get an overall sense of 
the problematic nature of an item encountered in the text of the poem all that 
they need. An option exists at the bottom of the “summarized entry” to access 
the ”long entry”; doing so, one finds that it includes far more information, 
including, on occasion, some bibliography, should the user want to have such. In 
my opinion, the Toynbee material is some of the most valuable available in the 
PDP, and I remain thankful to Oxford University Press for granting us permission 
to include this material in the database. 

If this Toynbee option is perhaps the quickest and most helpful way of 
bracketing the elements of a certain kind of query, it hardly exhausts what is 
available within the PDP. If one wants to know what the person who designed 



Humanist Studies & the Digital Age  Hollander 

3.1 Spring 2013  59 

the project thinks about a particular passage, one clicks on the “Commentary” 
portion of the menu for the verse and gets a good deal more information. 
However, one obvious problem that remains is that the user now has the views 
of only two individuals. What if he or she really wants to deal with centuries’ 
worth of disagreement over this, one of the most contention-causing loci in the 
poem?  Here’s where the PDP can be most helpful, taking the user to all 
responses found over the past six hundred and ninety years in a commentary 
tradition that is simply the richest ever to have grown alongside the text of a 
single poem. The fifty-four of the seventy-five glossators (some of whom did not 
finish their commentaries while others may have ducked this difficult passage) 
collected in the Dartmouth Dante Project (of the perhaps two hundred 
commentaries that exist, some completed, others abandoned, in the annals of 
Dante scholarship) do not, among themselves, guarantee that the Dantist who 
consults them will come away no more confused than when entering the selva 
oscura of this difficult poem. What they do unfailingly reveal is the difficulty and 
attractiveness of this poem. It has had a hold on readers for nearly eight 
centuries now -- and there is no sign that its attractiveness is waning. On the 
contrary, the extraordinary success of the actor Roberto Benigni, not only in 
Italy, but in the rest of Europe and in North and South America, who has been 
bringing Dante to a wider audience in his road show “Tutto Dante,” tells us 
that, in a time not noted for its appreciation of great literature, Dante’s success 
is wider than ever. To say that the Commedia is the greatest single poem ever 
written does not represent an obviously untrue claim; in any case, there are 
probably less than a handful of other plausible contestants for such a claim. 
Suffice it to say that the Commedia has enriched many a life and that the 
odds very strongly favor its continuing to do so. 

     It may be helpful to add that, if a user has questions or suggestions, 
there is a form for communicating with us on the PDP welcome screen (one 
simply clicks on the legend “Bugs? Comments?”). In the twelve years of our 
existence we have had quite a few of these, some pointing out typographical 
and other mechanical errors (assistance that is most welcome), others asking for 
information not available within the parameters of the database. While a few of 
these questions have not been answerable (at least not by me), I have 
responded immediately to every one of them and promise to do so in the 
future. 

 


