Digitizing Arabic: A Story of Script Technologies | J.R. Osborn

all right welcome everyone I'm minu malim a professor of gender and women's studies and director of media studies for those of you who are not familiar with our program media studies is one of the largest Majors campus by and every year we organize a lecture series and this is the SEC sixth event we organize this semester under the rubricov media mediation and mediatization before introducing our speaker yes before introducing our speaker I would like to thank the center for Middle Eastern studies and the center for New Media for co-sponsoring this event I also wish to thank Chelsea Prieto and Nathan Espanish for their help in organizing this event I also wish to thank for helping us with this recording and this session so without further Ado I would like to introduce our speaker but his speaker is Professor J.R Osborne who is a scholar and experimentalist of communication his work explores media history semiotics communication Technologies and design Aesthetics with a regional focus of the Middle East and Africa Dr Osborne is an assistant professor of communication culture and technology and co-director of the alteration Lab at Georgetown University his book Letters of light Arabic script in calligraphy print and digital design was published in 2017 by Harvard University press and it follows the book follows the story of Arabic script and Technology from the Advent of calligraphic tradition through the implementation of the Unicode standard letters of light was also the joint winner of the 2018 British Kuwait friendship society book prize and received an honorable mention for the 2018 Albert Ronnie book award from the Middle East studies Association Dr Osborne also produced co-directed edited the feature documentary glitter Ross finding art in Dubai in 2012 and he quoted African art reframe dialogues and Reflections on museum culture in 2020. his current research involves cross-cultural and competitive analysis of diagrams and information visualization let me welcome Dr Osborne and invite him to start his presentation it's yours thank you very much uh menu uh thank you for for having me my name is Jerry Osborne um I'm an associate professor in the communication culture and technology program at Georgetown University uh it's been introduced by master's program that looks at uh digital technology and all its ramifications so if there are any undergraduates interested in Graduate Studies feel free to to reach out after the talk um it's a honor to be here thank you very much to Professor Lim and the media studies program for inviting me thank you to executive director Nathan Spanish and the center for Middle Eastern studies um thanks to Chelsea Prieto Neil um Gali and Shiraz Ali for publicity and administrative support and thank you to the Center for New Media and um uh UC Berkeley itself my father is a Berkeley alum in 1960 and I have family in the East Bay so it's always nice to return to the Bay Area even if it's just just virtually uh so thank thank you again um I'm going to start sharing my screen now um let me know if there are any issues um and my talk today is digitizing Arabic the story of script Technologies and we will get to digital Arabic but we're going to take kind of a long and Winding Road through history uh to get there um and I'll apologize also in advance for kind of painting with some very broad brush strokes but hopefully that'll be justified by the time we reach the the end the end of the talk um the talk is based upon the my book Letters of light Arabic script and calligraphy print and digital design that the new mentioned which came out in 2017 as well as research I've done um done uh since then and this is a comparative study of of Arabic script over a very long duration as she said um and one of the things that I'm going to be focusing on this talk in terms of Arabic script is the distinction between uh two different models of understanding how Arabic operates as a script and the first I'll introduce is the four-fold model um of horizontal segmentation which is how we often learn about Arabic now that it is uh Arabic letters since they're cursive have different shapes whether they're isolated or if they're the initial medial or final position in a letter and those shapes changed depending on what letters they're connecting to and I call this the four-fold model because there's four forms for each each letter and it's based upon those what they change based on how they are horizontally in a line and I'm going to be contrasting that today with another model uh which I call see if I get this we call the air the seven layer model which I borrow from a scholar named Thomas Milo which looks at Arabic not simply as a horizontal line but as a two-dimensional space with both horizontal lines as well as vertical alignments so in in this model it begins with the the Russian the basic forms of the letters and then these these get modified above and below first with the nocta that appear both above and below the main line then additional characters such as tashkil and vocalization and in it if you go through the calligraphic practice even there are specific rules on how to fill empty spaces to protect erroneous readings especially of things such as uh the the Quran so it's it's a model that doesn't just look at Arabic as a line of horizontal line of writing but is one that is a horizontal line with things happening above and below it it's a two-dimensional spatial idea of writing and I'll kind of show here the two models side by side and I'll kind of outline how that kind of unfolds across a different set of Technologies um The Talk itself historically is based upon letters of light and um kind of will go through three what I call technical breaks um I have the here the the HTML code for a break is a little bit of a a visual pun but the letters of light is kind of a thousand year history of Arabic script it's a very long long durray history as I said on that sense it paints with a certain broad stroke um but there it's broken up by these moments of technological change in the 10th Century the formalization of proportion script which we now come to understand is calligraphy then in the 15th century movable type printing comes which leads to all sorts of changes in Europe and around the world and then in the 20th century uh digital design and the reason I wanted to write uh this book originally are letters of light was when I was studying history of communication we read a lot of long duration histories of communication and they often were very eurocentric they kind of had this story that the album that gets developed the Greeks come up with vowels eventually Gutenberg starts printing in Germany in the 20th century the Americans and and the British come up with Computing and now we live in a globalized international world and um not only is that focused on on a European context it also prioritizes Technologies designed for the Latin alphabet the the alphabet that are used for European languages and I wanted to look at a similar duration with an in a different setting and so I looked at Arabic script to see if different questions arise when trying to do is this kind of global long-term uh um idea um history of communication in another setting and so those are the three breaks we're going to be looking at so with that uh we'll we'll jump in um the first is proportion script this is um what we now know is calligraphy it's it's known as which was developed initially as Legend goes by the vizier IBN mukla who has quite a uh a colorful life which I'm happy to talk about um later if we if we wish but the idea of proportion script is that the script is designed by proportions across the letters and those proportions are based upon the nocta the note the point of hitting the read pen to the paper and then with the note you can measure the alif in a certain number of nocta here at seven tall then the elite becomes the diameter of a circle and then all the other letters can be formed by using measurements of note in that that line in the circle for those familiar or who or who practice in graphic design this is the basics of graphic design right here it is point line plane but again again as a two-dimensional thinking of letters as these two-dimensional Creations um this is still used in calligraphic practice here's a fairly recent introductory calligraphy manual um that was from the British library and you can see the letters and different aspects of it marked out with Nook did a show the proportions of the script here's another example um from the uh suleimania crypto fantasy um archives in Istanbul in the ottoman era and these red lines for those familiar with um Eastern Arabic numerals you can see these red lines have um numerals above them fives or threes or or fours that show what the various measurements are along these tangents and um measuring these letters based upon the the size of the tangents actually isn't too different from certain aspects of digital design when it's designed with busier curves now where we can stretch these to make the different different shapes and widths so it's not too different already but we're jumping ahead of ourselves a little bit what did is that it's standardized writing based upon these defined proportions and here's an example of a of a part of a Quran written by IBN in the early 11th century um I apologize also all the dates in the this presentation are in Common Era or the the the Christian calendar um I didn't also have them in in anahijui uh just for consistency's sake um but this is mainly in the 11th century and what's amazing about it is that now it remains easily legible to those familiar and literate in Arabic reading ten centuries after it was handwritten that's an amazing technological feat if we think of trying to read English handwriting from even a hundred years ago it's very difficult sometimes I can't even read my own scribblings that I did you know 20 30 years ago um when when I was taking them for myself so the ability to have a standardized script that was handwritten last this long is a wonderful technological feat the other thing does is that it makes different scripts identifiable based upon changing the relations between those proportions so here's the basmala written in a few Styles and depending on the proportions and and things whether you fill in letters or make them the terminals pointing a curve you get different styles of scripts that become visually noticeable so here are three scripts and here for example if we look at the meme sometimes it's filled in sometimes it's open but uh sometimes it connects with that above sometimes it comes to a sharp terminal these are distinctive features that can be noticed by the literate eye as different styles of script um here's that same the same phrase and again um with more more varied kind of a nostalic taliq style and a Muay diwani style and again these are those shapes I showed they they vary quite dramatically depending upon the proportions that are that are used to establish to establish them here's this the same idea this is a phrase um that was written by Master calligraphy Mohamed Zakaria to show the variation across these scripts um uh which he graciously um he graciously contributed for for my my book these are Primal Sita the um the six classical pens which are often paired this way and they have a um an air of kind of classical and austerity and by the time we get to the ottoman period so I'm going to jump forward a little bit now my first kind of historical jump these are very formalized as cert for certain types of content um the one I'll point out I hear mohawkick for example was is strongly hierarchic it's used often for headings in in Grand qurans or or kind of um religious um plackets and display pieces it's never used for example for uh Prozac or secular scientific text Nash is as a copious hand is more generally used so these they have very different different um types of uses depending on the script in ottoman practice um I have a recent article more recent article that came out last year that looks at this system of autumn script variation as well I'm happy to talk about that later here's a few more variations um and this again an ottoman practice talika diwani siaka these again had different different um uses according to genre and the the the the text that was being drafted um here's the whole collection together and just to point out some of that variation of the same phrase how different it looks these are the classical scripts some of which like mahakuk are highly hierarchic they have this kind of sacred character these are my what we might call a bit more secular scripts for the Ottomans taliq was used for poetry especially Persian influence poetry it was also used for religious decisions of the judges which would kind of distinguish the decisions which were commentary from the religious texts which had a different script diwani was used primarily as the script of the Chancery of the royal of the royal Bureau for Royal pronouncements and things like patents so that we we knew it was coming from the religious or from the um from the uh Royal um it was a Royal Decree which is why it's quite fancy and and and difficult to write siac as a shorthand was used mostly by Financial registers so these were used for different genres of text and as a mode of information then management you could look at the scripts and kind of get a sense of the genre maybe the origin of text who it was addressed to um and the metaphor I use when I describe this for example to um to uh my students or to others is that now we often organize our information spatially we have you know even in newspapers on a web page we've got the sports section the business section the education or entertainment section and they're all organized spatially imagine if instead those were all kind of combined together maybe because paper wasn't a premium we didn't have enough space but they're all combined together and instead they were defined by scripts that entertainment news was always in Comic Sans and and um Sports was always in helvetica and current events perhaps in business might be in in something austere like Times New Roman and they're all there together but then if I wanted to know for example who won the Warriors versus Kings NBA playoff game all I had to do was look for helvetica bolt that's that Sports and I could find it so it's a different way of organizing information a visual organization and again this would these are practices that developed long before printing arrived I'm going to um quote our historian Oleg rabar from his book mediation of ornament who talks about um um uh as sub as a technological Discovery did not consist in the establishment of what has now been called calligraphy of writing for its own sake and for aesthetic purposes it was possibly much more like the typographic inventions of the 16th and 17th centuries in Western Europe what was created was a series of types which were both both in the scientific and in the printing sense of the word a new standard was elaborated that allowed for variations and noticeable variations as any type would and the proportions dividing each script gave the process a mechanical side which as in printing guaranteed quality and consistency both across time and space and again this happened before for print it's standardized writing in a particular way and developed certain modes of writing and reading practice so then move forward a few hundred years into print when print arrives and the ottomans are using this this practice this was developed quite um quite um robust ottoman practice I would argue well what happens when this comes when print begins when I was studying again history of communication this is the story of print that has often given regarding the Middle East Gutenberg printed his Bible in 1455 Ottomans did not Grant ibrahimovertica the first ottoman printer didn't print until 1727 so there's about a 250 year delay in the in an ottoman practice and this you know can is used to explain a lot of things but the focus is if they're somehow a delay um I want to complicate this story a little bit first of all by just introducing this date which the Ottomans take Istanbul the conquer Constantinople in 1453 a mere two years before um before um Gutenberg printed his Bible um and during these early years of printing the Ottomans were in ascendance their culture was Advanced they were um they were on the rise they were spreading rapidly into Europe um and it was unified the the the the the land was unified in a single Administration whereas Europe was much more splintered Europe after shortly after Gutenberg was about to lose its religious unity through the Protestant Reformation there was a lot of competing administrations print actually spread through a lot of these kind of piracy and lack of administration so there was a very different socio-cultural settings um the other thing I want to stress is that print was not unknown in ottoman lands um Sultan mahmet uh who who conquered Istanbul was said to have printed Works in his Library um the when Jews were expelled from Spain and Portugal many of them went to uh Istanbul or Constantinople and brought um printing materials with them uh the first um print Hebrew printing in Istanbul was done in 1493 um that's less than 50 years after Gutenberg the first Torah was printed there in 1504 there's a vibrant Hebrew printing Community later they were vibrant our medium printing communities there was Greek greeting communities there was not but there was not printing in in Arabic script um there are rumors that the reason why is that The Sultans banned Printing and Arabic punishable by death um this has not been verified and there's a wonderful article by uh uh late colleague Catherine Schwartz who unfortunately passed away tragically uh last year called did the Ottoman Sultans band print where she looks at the historical record and shows that these bands are mostly began with European rumors they're not there in the archival tribal record so printing was known among non-um Arabic script communities and was being done successfully and and very and and well done in in ottoman lands number two the other example is that well Arabic was a cursive script that was difficult to print to align everything together to get the letters to fit this was also a technical feat that was solved by um the end of the uh 16th century these are two types by Robert granjan which he'd cut in in Italy for the typographic medicia um but these types were were made by this horizontal segmentation of of the the letters so these are very cursive they fit together they're they're highly legible to Arabic readers um but it's a different technological practice this is what a American type case looks like for those unfamiliar with it the movable type pieces would go into each of these boxes and you'd pick them out to make the letters by stringing them together this is not too different from how we now look at the four-fold model of Arabic script where there are different shapes in different boxes um here is a a map of a type case for uh what a for an Arabic type case might look and we can see for example the iron and its different variations down here in different boxes where they would keep these different pieces of metal movable type and to make a text you would take these pieces of individual metal type these little our nuns up here and you'd slide them together in a chase like this in lines and then that would form the form which could then be printed over an over and get multiple multiple copies um when we look at Grand Junction's type um we can see this um in in in practice if we put these lines here um these are what we call X Heights in Latin typography they're based upon the lower case X we can see that all the letters are in a standard Baseline they all kind of fit within this other line some go above and below and then if we put these vertical lines we can see everything fits in its nice little box maybe you can you can see that if if it's not too small but everything fits in its nice little box and it you can just slide them one after the other in in line and make a a line a line of text um if we're going to compare that to this phrase I showed before and this is a bit of a generalization because of the stylist nature of the Bismillah but just to as a to to to make the point if we were to put these lines I don't even know where to put the horizontal lines there are some Paws that are way above it things go below the letters if we're going to try to put vertical lines we're cutting through things you you how to even arrange that type is is it complete uh it's completely different style it's based upon the idea again of writing being a two-dimensional space rather than this single line strung strung together uh so what why is this important well European printers adapted Arabic script to the technology of movable type printing although this ease type setting it reduced the bounce and visual variation of handwritten script and this in turn altered the expected Rhythm which made printed letters difficult to read for those familiar with Arabic and to demonstrate that if you find this kind of all caps difficult to read it's because that also disrupts the visual Rhythm we're used to seen and so I think this standardizing along the line made a printed Arabic look very unfamiliar and it made it difficult to read more tiring to read because it had a very different visual um visual rhythm so when we when the Ottomans do begin to print when ibrahimovica prints in um 1726 we see a lot of efforts to make it look more and to provide some visual bounce that's much closer to handwritten um handwritten Arabic at the time handwritten not script at the time um here we compare it to Grand Junctions we can see how this these are very straight lines and this they have a bit more of a bounce we have certain features here you can see the fee is two things slid together in Grand Junctions but the fee in mutiferacas is a ligature they stacked together as one form muta verica actually had 500 plus different forms for his his set of type to make it look like handwriting this um this stylized invocation up here is actually all printed with one large carved printed form so there was this attempt to make it look more like printed um handwritten nask at the time the other thing that's very starts printing the Ottomans begin printing is what they printed the first thing the Ottomans officially printed was not a book it was actually a map um move to America printed a map and then after that what he printed were things like dictionaries official histories Geographic tomes a book on magnetism some of his uh his introductions have the first mentions of of scientists such as um Copernicus and Kepler that were that were ever um circulated in Middle East or middle eastern Islamic settings um so it's a different type of text these are all official and secular documents what Chandra Mukherjee calls material Goods they're there to help the administrative State he did not print things like classical texts or religious texts the Quran was not printed until much later in in ottoman um in ottoman lands and so one argument that I would say is print what it does is it doesn't just introduce well a new technology when it's adopted it's also a new style of script that fits into script variation that these new types of text administrative texts that are up-to-date Maps recent science are seen as a different type of genre of text therefore they have a different look printed letters that's that's consistent with the the practice um so if we go back to our our timeline we I want to introduce then a few other points at the end just before the ottoman's adopted print the first is in 1699 uh the French developed was called wil mandurah which is a um a specialized typeface that has idiosyncrasities on it and is only used by the national crest of France and that those idiosyncrasies allow the you to identify what is coming from that press versus another press because at the time there was a lot of plagiarism a lot of different presses were printed in different things and there was a difficult knowing where things originated so the the French made a very specific typeface that they could identify as things coming from the in Premier National the other thing that happens in 1710 in um let me go to Giles the statute of Anne is developed which I won't go into but is often seen as the precursor ring What We Now understand is copyright law protecting authors from things like plagiarism and unlicensed duplication and printings so this is a shift which of Korean historian Carla Hess has said changes print from a means of mechanical reproduction which is fairly unregulated in these early centuries to a mode of cultural production in which she calls the modern literary system where there are administrative constraints and limits to understand and control the information so it's it we know where it's coming from authors are protected printers are protected and that's a different type of cultural practice and so if we that's when the state starts to administer print and if we look at the ottoman adoption of print in line with these changes the ottomans are fully in line with the rest of Europe they're adopting it for State purposes in the in the early 1700s just as the other parts of Europe what they didn't do is they didn't adopt it dramatically when it was still in this kind of experimental unregulated stage and again I would say what queen did then at least as mutaferica understood it was that it fit into ottoman script variation new types of text this time this secular kind of text of scientific knowledge that weren't available were printed with yet another different looking script which would differentiate them from other manuscripts and texts already on the market and we know script variation remained a factor because one of the things that that moved to ferrica did print was um a this grammar Turk which was a from uh uh the for the French Consulate by by Holderman it was printed for French Consular staff it was the only thing um muchafferica printed with Latin characters as a phrase book and a basic Turkish grammar for French consulary staff and the table he gives of the letters in this book is in grave table I don't know if you can see it but rather than having a four form model what it has it has the letters of the of the the Turkish Arabic alphabet and instead lists shows those in a variety of different scripts which are um including um sulus or tulip as um as as it said in Arabic um yakuti which is a style of mohaka nest um and and kirimo so it shows these in different styles and there's a note on the bottom here that says the Turks have many other styles of writing which were emitted in the interest of space and if you read the accompanying text the the introduction that moved to Fairchild rights explains how each of these Styles has a distinct purpose used to quote Nest is the preferred script for copying the Quran diwani is the script of the royal Bureau taliq was used by judges and Poets karmas for record-keeping Solas for titles of books and patents and this is very important that these notes were in a book directed towards French Consular staff that they should be aware of these distinctions because it was diplomatically important to know if they were looking at a piece in Arabic script and they were just learning Turkish if it came from a religious judge versus the Royal Bureau whether it was a patent or what whether it was some type of record keeping these practices of script variation had administrative purpose and I think this is evidence in that regard so nevertheless multiferica does print there's a bit of a dip after mutpherica Prints but then by the late 1800s in the 19th century 1800s print culture is developing in um ottoman lands and across um the the Middle East more more broadly nevertheless um uh these rumors that Arabic somehow doesn't work well with print technology persist well into the 20th century that is somehow backwards it doesn't quite fit its cursive structure is too flourishing they all these these these keep going in the 20th century and I'll give just three anecdotes about that the first um in 1928 um uh Mustafa come out of Turk when he unveils the new Turkish alphabet shows that these are now Latin based letters they they forego the Arabic and Autumn ottoman past an Arabic script as a break with that past but also one of the arguments was that the Arabic script um of Ottoman Turkish was not could not be facilitated with modern Technologies they could bring in from Europe so they might very much a line that wanted to align them so you're both technologically as well as kind of ideologically and Visually um a few years later um the English interpretation of the Quran one of the became one of the early um inter English interpretations that became very popular in a circulated very widely by Abdullah Yusuf Ali has the English text set in movable type while it preserves the Arabic script right next to it is actually at this time photo um photolithographically produced in order to preserve the handwriting of the calligrapher pure Abdul Hamid so even in this translation this interpretation of the Quran the the English text is typeset but the idea that Arabic should remain kind of more as an image with this handwritten character and it does not come into the into the technological register of print and then finally and most dramatically um in the middle of the 20th century The Academy of Arabic language and Cairo sponsors a contest for Arabic scripture form seeking Solutions on how to um how to simplify or reform Arabic script for modern Technologies um here's just one of the most drastic and dramatic proposals um where Arabic letters are played with one form reminiscent and kind of Drawn Together from Arabic or from Latin characters um this might be a little familiar to those who who know arabisi as a more recent phenomenon um ultimately none of these these um proposals were accepted broadly um recently uh last year I I published a translation of a a text by one of the judges on this committee who also had his own suggestions of reform uh with my colleagues Catherine Schwartz and Mohammed Jabar talking about this reform movement um but the movement was was was there thinking that Arabic again still needed to somehow change to fit in with the technology of the time none of them were adopted um and I will say as well that this wasn't unique to Arabic there's this the middle of the 20th century you see a lot of things Chinese opinion develops there's attempts to simplify Hebrew there's attempts to simplify other Chinese Guardians uh Indian scripts there's a lot of different attempts to um kind of fold things into the technology that was designed for Latin Latin technology but none of these were adopted for Arabic which finally brings us to digital and text and brings us to this question I probably now have a think of about 15 minutes more to go or so um brings us this question of digital text and Computing are we going to continue to design Arabic for technology to make it on this flat line by putting it together which is very looks very much like a Latin line of letters or are we going to somehow make technology that is designed for Arabic this is is actually um a very stylized necelic phrase um in nostaliq just to emphasize the point again of of this variety of nastoleek being very two-dimensional spatial things above and below each other rather than a a line it's there for kind of um for effect um the year nordsage the typographer describes typography as writing with pre-fabricated letters but in fact through most of the past 560 years since Gutenberg typography has been writing with prefabricated rectangles that fit into these boxes this is the Gutenberg Paradigm the arrangement of lines of rectangles that gets slid together one after the other to preserve a consistent line that then form images these form images of letters that then come together they're plotted against each other form um word images is and it's a very simple and very effective um solution and the Simplicity is allowed it to thrive but it's based upon the idea that these slide things together in static lines um and that each letter has its own box this has been carried over into the digital realm here's the Unicode standards basic Arabic plane and again each letter gets its box with its individual code points leading from top and bottom here the the various characters now computers can have contextual variation so you can have it you you get a code point from Unicode it looks it up it brings back in an image of that letter then if you look up another code point it brings back the other letter if it's if it and because computers can work with variation it knows that well if it's now goes from an isolated to beginning it's going to change shape the next letter if it's going to be middle it's going to get a different image but it's still looking using these characters to look up images getting that image and putting them in a row Side by by side how this works um with a Latin example here is that all these letters get lined up this is a slide from John Hudson um and then you you verify you make sure they're readable you see some problems here with ethani kind of squeezed together there's a bit too much space by the V and the E so then you just kind of shift the boxes you put make a ligature put FDI together maybe slide this box a little bit under there but it all fits together and there's still boxes so what do you do in the Arabic case when the digital technology is based upon these boxes for Latin letters well here's one drastic solution you take Latin letters as they are which fit into very nice boxes that's what they're designed for the technology designed for those boxes and you start cutting them up you maybe kind of take the top off of a g you flip over the M you slide them together you rotate another L you chop off all the parts of letters you put on a sideways colon and you end up with something that ripfully resembles an Arabic word if you're a solid written message it's readable it's not very pretty but it's there this is comparing that with um Arabic Giza Pro from Adobe um and the letters are there you can see it's it's legible they both they all fit and they're very nice boxes so they can all fit on a line and be slid together uh this was a technique in early Arabic font design to get the visual variety of of Latin fonts into Arabic Ahmad humid who I interviewed in the early uh 2000s did this he hacked um what's known as the Ahmad font from helvetica Nero he took a Vatican air chopped it all up and made it into a an Arabic font he was the editor of bite Middle East at the time the computer magazine and wanted a consistent font for the for both the headings of the Arabic as well as the English headlines um so but it was completely made by helvetica by cutting already had that visual consistency he was looking for and he could use we use those those those forms this remains big business in terms of branding for those who who can read this here's the Giorgio Armani logo a mall in Dubai and we can see the I of the Latin letter is the same as the elite so it's still big branding in terms of visual consistency in a in a globalized world of getting these Latin characters to look the same in different contexts um you can do a whole branding exercise looking at at um these um these adoptions of of letters both ways um to to maintain visual consistency and then there are Grand projects such as typographic matchmaking which was organized by uh Huda Abba Ferris um in which the which Dutch design Dutch type type designers were paired with Middle Eastern designers to make uh matching Arabic fonts for for French or I'm sorry for Dutch typefaces so this was the Fresco typeface this is the Arabic companion and there's all these designs to see to get the terminals on top to have a similar look and feel um as well as the um as well as the the bottom Terminals and finials to hear line spacing is preserved it preserves line spacing in terms of leading as well as Heights so if you print them and go between the two languages it's not gonna just just not gonna disturb the lines they're going to be consistent they can they can work side by side and this is a wonderful opportunity this opens up a great um um kind of cross-cultural dialogue and allows these languages to be side by side to have multilingual text for Arabic text to go into venues which they couldn't um go into digitally before because they now can fit in these Latin grids but there's still a certain direction going the companions always traveled in One Direction in this project they always were Dutch fonts which then got an Arabic companion it never went vice versa and there's a tacit assumption that's starting with Latin typography would remain the Baseline for international design it standardizes things along what the what Latin typography was doing but there's another way possible and that brings us back to this question um that I began the digital section with um and this is a question from Thomas Milo who who introduced the seven layer model I I showed earlier and Milo took a different approach and designed with team at Deco type Cas meme which is a a plug-in for InDesign um and runs on the advanced composition engine this that we're looking at here with all its calligraphic kind of extensions and the layering of things below these swashes that go below letters is still all Unicode readable and searchable even though it looks very calligraphic it's designed in in in in TAS Meme and has a very different idea of font rendering than most of our our Technologies in short what um the advanced composition engine does in Tuscany lets you do is to take a line of text like this which is very straight and make it look like this very calligraphic very layered and if you're interested in watching that you can um you can uh go into YouTube and and put in Tasmanian 4.0 and watch the the watch

the the change from one to the other I'm not going to show you that video I am going to show a slightly um different video um I uh shared so he did a video just to show the variations that are possible in this system so let me see if I can change to my video player I hope you should now be seeing the video player and I'm going to start it playing this is um the word ten bihi him which is just used because of how long it is and it's going to go through all the variations from the shortest to the longest possible variation of this word that the test meme can support um I'm going to start it playing so we can see it going um and you'll see so there's variations on the hair there's different things on how the kashida and how long words can be extended you can speed it up even a little bit and you can see it cycling through all these variations on the the letters um can go from incredibly kind of more compact this is the most compact one on top to to much longer depending on your spacing you can see some of them are layered up above some of them all on more on one line there's an incredible variety um the number of variations here is in its thousands um but not all the variations of connections are allowed um so y'all all all maybe pause it so you can see some you know how you can determine how many kashida how long you want your extensions the number of variations of thousands but not all of them are allowed all of the connections that are shown in this model are based upon analysis of the handwriting of a certain scribe muzette is that offendy and and all of these handlings appear within this Corpus as specific connect calligraphic connections or scribal or Kato Monsoon connections that he wrote and and those and so these aren't just randomly generated they're all based upon this Corpus to to give options that are that are both scribal and um and Visually pleasing the technology therefore let me go back to my um other screen if I can let's see um the technology there for hopefully you're seeing the the slides again um both remediates not just the letters of Arabic but also the visual style and um Milo received the prestigious Dr Peter Carroll award for plant technology and digital typography through this sophistication and promise of this of this engine so um so just do I want to try to close something really quick on my screen so how does this work well rather than using this horizontal kind of racketing of letters one after the other develop the seven layer model where you have layers of script and when you add another layer things get rearranged you can put you can put things such as nokta here between the Baseline and other letters you're thinking of it more in a two-dimensional layout um oops go forward again not just this model of four forms scripts being slid one after the other here's a comparison of um on the bottom Adobe Arabic mask which we see again in these ice boxes all in one line to how those boxes would look in the advanced composition engine from Deco type again they're layered on the boxes are layered on top of each other in this model you can't you can't layer them you can only just slide them together they're layered on top some of them are higher they have different bass lines it's a two-dimensional space of writing um here's a much more dramatic example this is not Arabic this is in in Urdu um this is set in the uh Deco type nastelic typeface designed by Miriam Summers um and so it's not just a change in plot but it's an entirely different idea of font rendering if we were this is also example from John Hudson if we were to go back who gave us those verification boxes I showed earlier and if we're going to put boxes here they'd look kind of like this we take away the letters these are what the boxes would look like for traditional font rendering to make this what we just saw incredibly complex um but the thing about Deco type and testament is that it doesn't use boxes and instead it uses Strokes it drawing commands to connect letters and strokes can be layered and and put together it doesn't think about boxes with an in a letter image in it it thinks about the commands to draw the outline of the letter um so it uses Strokes which makes this incredibly complex visually beautiful piece of piece of writing possible um it still relies on Unicode and the Unicode codes but what it does is it renders Unicode differently it doesn't look up the Unicode code and say okay here's here's the code Point here's the code Point okay that gets us to to fit put that in okay look up the next code Point find the image Slide the next image together in boxes it doesn't use these code points just to look up images and put them in which is what the kind of the digital version of movable type square box square box slide them together instead it uses these Digi these these Unicode code points as code and it uses that code and it says okay this is going to draw this letter now we're going to program the stroke of that letter whether it overlaps with something what the outline looks like and the variations that stroke can have so this is the last uh slide in this this regard um here's where to him um and each of the letters here is a different color the Ted The Head and the meme um and we can see the different ways that the stroke can be drawn um depending on if you want to have it kind of going from top to bottom if you want to have it more kind of on a line some of these connections The Strokes can overlap vertically and connect vertically the gene can connect vertically here you can have different different variations of the the letter H those are different commands to draw a Different Stroke the or they can connect again they connect vertically or horizontally you have these different parameters of how it will be drawn that you can set and then Unicode will look up the codes then depending on these parameters well then draw the letters much more of a two-dimensional space um so I'll wrap up wrap up there and I'll just say that the letters of light what it did in the experiment of looking at uh Arabic across this long duration and a non-european case of a Thousand-Year script history is that all their patterns emerged in this case there's a tradition in Arabic writing to see writing as a two-dimensional practice of design with horizontal and vertical components not just a unified single horizontal line of writing and that like that that led to different developments across all of these Technologies in terms of the development of alcata on sub adapting it to print and still hear ramifications in terms of digital design right now and then the the final argument I'll say in this type of study I think for media history is that looking at these types of studies not only give us a comparative idea of media history and a better idea of the various paths we have taken to get to our current globalized Internet Internet world of multiple scripts and multiple fonts but they also can suggest lost practices and traditions and new Vistas and exciting future imaginaries for digital design I think the advanced composition engine has enormous potential not only for Arabic but for Latin script as well and all sorts of varieties which I'm happy to to discuss and suggest but studying these different Traditions not only gives us a better sense of the past but also can help us imagine different ways that we can communicate in our digital present and our digital future so with that say thank you shukran and I'd very much look forward to continuing um the discussion with your questions and answers and comments thank you very much so much Dr Osborne for a very informative and wonderful talk and at this point you can either post your questions on Q a or raise your hand and ask your questions so uh let's see so far we don't have any question um however there was a question about accessing this talk uh in the future I think this talk is being recorded and you should have access to the recording you should contact the center for Middle Eastern studies and get them to share the recording with you so as I have before if there are no questions let me ask you a question my final slide I'm happy for those who have questions yet and want to email me after for certain slides I'll put that up to my email at Georgetown is there as well thank you so much AR so let me ask you a question you mentioned that digitization has been very eurocentric and and you mentioned that digitizing Arabic might raise new questions or uh new ideas so could you say a little more about some of the specific questions that arise from digitizing Arabic also a follow-up question let's say there are many students undergraduate students who are studying media studies why should they know about digitizing another language you know in this case Arabic what is important about that if you could just say a few words about that of course it's you mentioned that you've had lots of ideas offered lots of ideas but if you could say something to the audience who might be undergraduate students in media studies and they're thinking about these issues or they might be undergraduate students in communication or Computer Sciences or journalism what is the importance of digitizing another language um well I would say that um yeah I I first I'll say that that the Unicode standard and International Communication now I think it's fairly multi-scriptual and multi-multi-linguistic and it works uh well in many regards um however there are still paradigms at play which I think implicitly draw from certain Traditions that we might not be aware of um I would say that the biggest lesson from from from Arabic is it was this difference that I I outlined of thinking of writing as kind of two-dimensional space um with kind of both horizontal and vertical registers rather than just a single line of text lines of text are very easy for digital texts because we transmit things as lines of zeros and ones right lines of text um were printed then into Telegraph when we were typing tapping things out you just had a single what is the the letter after letter so the lines of text are very well known um and of course we speak in in temporarily we speak in a single line so this idea that that language follows a certain line is is there and it it's a powerful idea and it's it's led to a lot of amazing developments however the history of writing is it's is language in two Dimensions it's not language Justin in one dimension like speech and there's a lot of there's been a lot of um different developments and I think one thing Arabic tradition teaches us is that the the oh scene design and graphic design now is a two-dimensional practice and seeing writing itself is that practice of design in with the difference with with vertical as well as horizontal changes really opens up the possibility of of communication we and we see that not only in Arabic we see this is what graphic design does whenever someone's designing a logo and trying to get you know Latin fonts to do something different and appear in a different different shape and so that I think is is it is a is a general practice of design that our technology can afford board um the question is what how can we how do we how can we produce those texts our technology also constrains the way we communicate nice the reason something like arabisi which is texting with Latin characters spread in the Arab Spring was because a lot of early cell phones had Latin characters but didn't support Arabic so you had to adapt things like a three for iron and different characters um like like like this to because you had because we communicate with the tools that are that are available to us and if we and if you and in the in the Arab Spring if you're limiting all you have is Latin characters you're gonna use Latin characters to um to to represent as best they can Arabic phrases because that's what you want to communicate and so the question is making our tools and Technology able to share the messages we want what I showed like Tas meme is a wonderful software and it can do all these incredibly calligraphic visually visual um things in in design in design we still understand as a very specialized software our um our um are most of the things we do our writing in something like Microsoft Word or Google Docs is still based upon this lookup model of Unicode where it's just getting a picture of an image and putting it next to each other and that works wonderfully but but if if there's a way we want to express something through a more fluid script that technology is not is not available to us so I think it's that question of opening up the possibility of communication and um by studying alternative traditions we become familiar with the different ways that that that writing has been practiced the different um and and things that may be lost or or um unwillingly lost in in terms certain types of digitization projects um yes great there are two great questions would you like to read them or do you want me to read them for you um okay um I can read them so the first one for uh have you surveyed other printed material other than muta ferricas um Aleppo so um I I don't know if uh you know if you have if you have other um comments on on on the on the the question related to printed material I have looked at other um pretty material my my focus in the in in the book and most of my focus in terms of printing is looking at mutiferica because I I the based on the book the research is kind of these three kind of breaks um but there have there there are a lot of um a lot of variations of course in in um in in Arabic printing um I believe the Aleppo one is looking at some of the earlier versions of um an earlier printing um that happened before moved to ferrica and it and it's true and I will say this right now was he was not the first to print Arabic he was not the first to print in the Middle East he was not the first Muslim printer he was none of those things he was the first officially sanctioned printer of the Ottoman Empire which is I think again the shift to State organization and support support of printing um so um so that that's that's just to say that there were printings be before him that and and and um in in the Middle East um the Aleppo print I believe it's the furniture was a book of prayers for the for the Christian Community um there's still this question um of why print um was kind of the print culture kind of developed in the 18th or 19th century rather than as quickly as it did in Europe and I think rather um then because even these earlier Prince even moved farica prints weren't in his best sellers Corinthians before him it didn't kind of create a tradition and um and my argument is rather than this technological disjuncture which is often the the presentation of this delay in ottom in practice it has to do with the the fact that the the the Ottomans I'd say the greater the greater Middle East and Arabic writing world was operating in a writing system that was familiar with visual variation in a different way and read those variations as signs of content and marker and printing early printing was very good at standardizing and visualizing standardizing text but it wasn't very good at visual variation so I think it takes a while for that visual variation to come back into into Arabic so I hope that that addresses some of the some of the the questions um not I'm happy to answer another question by harun um so for Harmons is um printed letters are now machine readable however do you think that handwritten text can be readable by Machine mean that it will be possible to get them to be ocr'd um so yes I think there's been a lot of um a lot of um Improvement on this I think the Arabic scripts both printed uh mechanical Arabic scripts as well as handwritten Arabic scripts early on um had trouble um OCR had trouble with them and I think OCR had optical character recognition making it making an image readable had trouble with um scripts because um with handwritten scripts both are I think Latin as well as Arabic in particular is because OCR early OCR was reverse engineering the Gutenberg Paradigm it was looking for a box and would find a shape within that box and says this shape is this letter and then it would look for the next box next to it the next box and if you're looking for those boxes on a line it's not going to work very well in a handwritten Arabic script because the line bounces too much so there have been a lot now OCR is starting to break that I think it's still kind of based upon trying to locate boxes but it's um but I think it's possible I think there's been a lot of um of developments in this um um um a um uh uh who's saying I forgetting his last last name currently at the center for manuscript studies where I just had my fellowship has a wonderful kind of handwritten analysis tool that you can feed it examples of images of handwriting and it can pair that handwriting on other with other um text it's best guests based upon reviewing the patterns so oh so I think the trick with OCR is that it's OCR is if it's looking for boxes with letters in it it's still it's looking for Latin Style a Latin idea of letters in a row and it's going to be tricky if it's looking for patterns it's much more open and and the example I'll give that's not related to text in this regard is that if somebody uploads my image to Facebook Facebook's algorithms will go through all these and based upon our faces it'll say this is Jerry Osborne this is my new this is various people because it's basing it on pattern recognition of a two-dimensional space right and so those Technologies to find these two-dimensional patterns are there and quite robust and it's getting that same type of thing and applying it to to script which you'll then understand okay writing isn't just in this line of boxes it's getting understanding these two-dimensional patterns between certain things sometimes the gene connects from the top sometimes it's from the side but it has these things that preserved or these types of proportions are always consistent if you once OCR starts looking for those patterns more broadly it the the it'll become um I would say become easier but it but it's a different type of of problem it's it's it's looking at the visual patterns and saying what are the visual within this visual written piece what are the patterns that there that can be identifiable and that's a different type of machine learning than saying within this piece we are looking for kind of boxes with letters within them which I think limited early Arabic OCR yes uh there are no other questions and oh no there are more questions okay uh no okay that's that's they are thanking you so let me ask you another questi

2023-05-03 07:09

Other news