Kis Cyrillic

Here is the second and fi­nal part of Vladi­mir Ye­fimov's es­say Civil Type and Kis Cyril­lic. Ac­cord­ing to the au­thor, the typefaces of Hun­gari­an geni­us Miklós Kis (1650–1702) are the most strik­ing ex­ample of Dutch old-style An­ti­qua. In this second part of the es­say, Vladi­mir Ye­fimov gives a de­tailed ac­count of how Cyril­lic script for the Bit­s­ream's Kis—a con­tem­por­ary ver­sion of Miklós Kis's typeface—was cre­ated. One of the most beau­ti­ful Lat­in typefaces of the great Rus­si­an ruler's era, in the au­thor’s opin­ion, gives the an­swer to a very reas­on­able ques­tion—Could a Pet­rine typeface take a dif­fer­ent form? The text is com­ple­men­ted with new il­lus­tra­tions and pre­vi­ously un­pub­lished draw­ings from Ye­fimov's per­son­al archive.

24 January 2014


Vladimir Yefimov


Maxim Zhukov,
Misha Beletsky

t is evid­ent that the design of the new Cyril­lic type was done at the am­a­teur level, al­though this was prob­ably the first at­tempt in Rus­sia to use a mod­ern ap­proach to type design, in­clud­ing sketch­ing, the cre­ation of ori­gin­al draw­ings, the man­u­fac­tur­ing of type punches and matrices, then the print­ing of tests, with suc­cess­ive proofread­ing and im­prove­ments to the type. We don’t know which of the ro­man faces was used as the start­ing point for the design of the new let­ters. It is most likely that there was no single pro­to­type and that the struc­ture of the Civil Type was not based on one par­tic­u­lar ro­man font. The sketches were a kind of ec­lect­ic mix­ture of the dif­fer­ent ro­man faces avail­able in Peter’s time (from late Dutch old-style ro­man to Ro­main du Roi). As a res­ult, the Civil Type be­came very dis­tinct­ive, but in fact its dis­tinc­tion was just the styl­ised trap­pings of the ro­man style, an im­it­a­tion of it.

There were a num­ber of reas­ons for this res­ult (haste in gen­er­al, the war, lack of skilled people), but the main one was the ab­sence of a uni­fied hand­writ­ing script that could have served as the basis for the de­vel­op­ment of a new type. An­oth­er was the am­a­teur ap­proach to the design. Of course, Peter the Great was a spe­cial­ist in a lot of areas; as Pushkin de­scribed him, he was an “…aca­dem­i­cian and hero, nav­ig­at­or and car­penter…” But he was not a type de­sign­er, even if he made the sketches of the new let­ters him­self. Oth­er spe­cial­ists (drafts­men, punch­cut­ters, print­ers) were sub­or­din­ate in their status to the au­gust de­sign­er. And he had a very dif­fi­cult task, try­ing to com­bine ele­ments that were so dif­fer­ent in style (the old poluustav let­ters and ro­man char­ac­ters); in the ab­sence of any stable hand­writ­ten form for these let­ters, the res­ult of such a com­bin­a­tion could be only a hodge­podge. Later modi­fic­a­tion of the Civil Type was rather haphaz­ard, done by tri­al and er­ror: dif­fer­ent forms of let­ters were tried out, and new char­ac­ter forms were sub­sti­tuted for the old ones; up­per­case let­ters re­peated the form of the lower­case ones, and vice versa; and stresses were in­tro­duced and then re­jec­ted again. It took over a hun­dred years more after that type re­form to sta­bil­ise Cyril­lic let­ter­forms.


View of Am­s­ter­dam. En­grav­ing. 1700.

On the oth­er hand, some rad­ic­als (es­pe­cially some Rus­si­an graph­ic de­sign­ers) con­tin­ue to be­lieve that Peter should have giv­en an or­der to all Rus­si­ans to switch to the Lat­in al­pha­bet, as he did with beards and smoking to­bacco. If this had happened, from the cul­tur­al point of view this coun­try would have ended up not so dif­fer­ent from the West, and we would not have met so many dif­fi­culties with in­ter­na­tion­al com­mu­nic­a­tions and the in­ter­net; we would have been able to use all the treas­ures of the Lat­in types ac­cu­mu­lated by West­ern civil­isa­tion. So we should im­me­di­ately cor­rect the em­per­or’s mis­take!

I do not sup­port this view­point, though it is quite evid­ent that the Lat­in al­pha­bet has had a longer evol­u­tion than the Cyril­lic, and its artist­ic vir­tues are in no doubt. But Rus­sia fol­lows its own his­tor­ic­al way, es­pe­cially in type de­vel­op­ment. Still, the res­ults of the Pet­rine type re­form could have been dif­fer­ent; and be­ing a type de­sign­er, I of­ten try to ima­gine what might have been done if they had bet­ter ma­ter­i­al to work with.

The Way It Goes in Rus­sia

Of course everything should have be­gun with choos­ing the best Lat­in pro­to­type for the new Cyril­lic al­pha­bet. Hav­ing ar­rived in Hol­land in 1697 as a mem­ber of the Great Em­bassy, Peter could still see the best samples of the Dutch old-style ro­man in use. At the be­gin­ning of the 18th cen­tury it was pos­sible to en­counter El­sevir type tra­di­tions in many coun­tries, in­clud­ing Hol­land, Ger­many, and Eng­land, which had a lot of print­ing houses stocked with the types of such mas­ters as Cris­tof­fel van Dijck, Bartho­lomeus Vo­skens, and Nich­olas Kis (Mik­lós Tótfalusi Kis).


Am­s­ter­dam edi­tion of the Bible. Open­ing page of the New Test­a­ment. Nich­olas Kis. Am­s­ter­dam, 1685.

Nich­olas Kis (1650–1702) was the old­est con­tem­por­ary of the Rus­si­an type re­former. Of Hun­gari­an ori­gin, he worked in Am­s­ter­dam learn­ing punch­cut­ting and ty­po­graph­ic skills from 1680 to 1689, in or­der to cre­ate his own type for pub­lish­ing the Hun­gari­an Bible. He had to cut a lot of typefaces for oth­er type foundries to fin­ance the Bible print­ing. After fin­ish­ing his mis­sion he went back to Transylvania, es­tab­lished a print­shop in the town of Ko­loszvár (now Cluj in Ro­mania), and util the end of his life pub­lished dif­fer­ent kinds of lit­er­at­ure in Lat­in and Hun­gari­an, try­ing to edu­cate his people. While he was still in Am­s­ter­dam, Kis be­came so fam­ous and in so much de­mand for his punch­cut­ting skills that he got or­ders not only from Dutch type founders but from ty­po­graph­ers in oth­er coun­tries. Kis’s types were found in Ger­many, Po­land, Sweden, Eng­land, and Italy.

I be­lieve that his types are the most beau­ti­ful of all ro­mans, and at the same time they are not that dis­tant from Peter the Great’s epoch. It would have been quite nat­ur­al to se­lect these types as a pro­to­type for the Civil Type.

It is note­worthy that Peter knew someone who was ac­quain­ted with Kis. This was Nic­o­l­aas Cor­neli­us Wit­sen, the May­or of Am­s­ter­dam, a trav­el­ler and geo­graph­er, and the pub­lish­er of books on ship­build­ing and Siberi­an geo­graphy, who re­ceived the Rus­si­an Great Em­bassy dur­ing their stay in Am­s­ter­dam in 1697. Only a few years be­fore this, Wit­sen had re­com­men­ded Kis to the Geor­gi­an Tsar Archil (who was stay­ing in Mo­scow at the time) as an ex­cel­lent punch­cut­ter. In 1687, at the re­quest of the Geor­gi­an Gov­ernor, Kis man­u­fac­tured sev­er­al Geor­gi­an types, samples of which were dis­covered in the Uni­versity Lib­rary in Uppsala, Sweden. (In ad­di­tion to Geor­gi­an, Kis made some Hebrew, Greek, and Ar­meni­an types, and also Syr­i­an, Samar­it­an, and Coptic types, though the fact that these last three were man­u­fac­tured is known only from one of Kis’s let­ters).


Nich­olas Kis type samples. Am­s­ter­dam, ca. 1686.

About fif­teen years ago I made my first at­tempt at fol­low­ing in Peter’s foot­steps when I made some sketches based on the let­ter­forms of the typeface known as Jan­son. The pro­to­type of this typeface was found in the Haas-Drug­ulin print­ing house in Leipzig at the be­gin­ning of the twen­ti­eth cen­tury, and at first it was mis­takenly at­trib­uted to Ant­on Jan­son (1620–1687), a Dutch crafts­man who worked in Leipzig. In 1954, however, after com­par­ing type spe­ci­mens, the Eng­lish ty­po­graph­ic his­tor­i­an Harry Carter and the Hun­gari­an wood en­graver György Buday came to the con­clu­sion that the so-called Jan­son types were in fact the work of Nich­olas Kis.

In 1996, after the Para­Type com­pany in Mo­scow signed an agree­ment with Bit­stream to li­cense their typefaces and de­vel­op Cyril­lic ver­sions of them, I was very glad to come back to this work; I based my re­newed ef­forts on Bit­stream Kis, which is very close to Jan­son. On the one hand, this gave me an op­por­tun­ity to make a mod­ern text type of the late Baroque Dutch ro­man style, which at that time had no ana­logues in Cyril­lic ty­po­graphy. On the oth­er hand, I wanted to make its let­ter­forms a bit ar­cha­ic, re­sem­bling the Pet­rine Civil Type, and thus in some way cor­rect it in the mod­ern age, by mak­ing the shape of the let­ters less con­tra­dict­ory.

From the very be­gin­ning it was clear that the Cyril­lic let­ters that cor­res­pon­ded to Lat­in ones should re­main the same as in the Lat­in al­pha­bet. I was not go­ing to re­peat Peter’s mis­take and in­vent a new form of А and oth­er Lat­in let­ters just be­cause I liked the shape cre­ated by Nich­olas Kis. The form of the Cyril­lic char­ac­ters that are struc­tur­ally close to Lat­in should also be defined by the Lat­in al­pha­bet, so as to stay in the frame­work of mod­ern tra­di­tions of Cyril­lic type design. The prob­lem was not only in de­fin­ing the forms of the Cyril­lic let­ters that have no ana­logues in the Lat­in al­pha­bet, which can have vari­ant forms, but also in de­cid­ing wheth­er to keep Cyril­lic К, к sim­il­ar to Lat­in K, k or to find an­oth­er form for the di­ag­on­als, more spe­cif­ic to Cyril­lic. The same prob­lem arises for Я, я re­l­at­ive to the Lat­in R.

It was also ne­ces­sary to design an it­al­ic, and that was es­sen­tially a com­pletely dif­fer­ent prob­lem. Cyril­lic it­al­ics ap­peared only in the post-Pet­rine peri­od, in the 1730s; and un­like European it­al­ics, the form of the first Rus­si­an it­al­ic let­ters was based not on hand­writ­ing but on en­graved in­scrip­tions in book titles, geo­graph­ic­al maps, and oth­er prin­ted mat­ter. Only at the end of the 18th and be­gin­ning of the 19th cen­tur­ies did the form of Cyril­lic lower­case it­al­ics be­gin to re­semble the Lat­in. Prob­ably this is due to the mass im­port of Didot types from France, and to the activ­it­ies of the Rus­si­an type man­u­fac­tur­ers Au­gust Se­men, Al­ex­an­der Pluchard, and George Re­vil­lon in par­tic­u­lar. That is why the struc­ture and ap­pear­ance of mod­ern Cyril­lic it­al­ics are much closer to the cor­res­pond­ing Lat­in than are the forms of Cyril­lic ro­man, most of which just re­peat the forms of the up­per­case let­ters. Con­sid­er­ing the fact that one of the com­mon forms of hand­writ­ing in Rus­si­an of the 19th and early 20th cen­tur­ies was a kind of con­nec­ted script writ­ten with a poin­ted pen (so-called ‘Eng­lish’ cal­li­graphy), one can con­clude that the struc­ture of Cyril­lic it­al­ics—no mat­ter how para­dox­ic­al this may sound—is more settled than the ro­man be­cause it de­veloped later.

As a rule, in the lower­case Cyril­lic it­al­ics some let­ters are the same as Lat­in ones; the struc­ture of many oth­ers is re­lated to Lat­in it­al­ic or Cyril­lic up­right let­ters, and all that needs de­fin­ing is a few de­tails, es­pe­cially the as­cend­ers and des­cend­ers. Only a few it­al­ic char­ac­ters re­quire spe­cial at­ten­tion, but their ana­logues can be found in the hand­writ­ten scripts of the 19th cen­tury. Up­per­case it­al­ic let­ters are usu­ally made by in­clin­ing the up­right let­ters and mak­ing any ne­ces­sary sub­sequent ad­just­ments.

Kis Cyril­lic

In the sum­mer of 1997, when I drew the sketches of the ro­man and it­al­ic Cyril­lic Kis on tra­cing pa­per, it was still not clear to me what changes ought to be made to the shapes of the Civil Type. In the ro­man, the out­lines of most of the let­ters that were spe­cif­ic to Cyril­lic were based on the types of the middle and late 18th cen­tury, and also on mod­ern Cyril­lic typefaces that in­ter­pret the forms of 18th cen­tury types (Academy, Ban­nikova, Eliza­beth). So in the be­gin­ning I was go­ing to make let­ters Жж, and К, к dif­fer­ent in nature, as they were in the Civil Type (Жж, к with drop-shaped up­per ter­min­als, and with wave-shaped legs in К, к). The double-bowled Ф, ф, the wave-shaped bot­tom ter­min­als in з, э, and the wave-shaped des­cend­ing ele­ments in Ц, ц, Щ, щ in gen­er­al re­peated the struc­ture of sim­il­ar let­ters from the Academy and Eliza­beth faces. The trapezoid­al Д, д, Л, л on the oth­er hand re­sembled mod­ern types (in Peter’s time the Д had the form of a right-angled tri­angle and the Л that of an iso­sceles tri­angle with in­clined side strokes). З, with a drop-shaped bot­tom ter­min­al, re­sembled forms of the Pet­rine en­grav­ings.

Fur­ther, in 1997–98 (on the com­puter), I changed the struc­ture of some char­ac­ters quite a bit from what I had done in the sketches. I re­defined the struc­ture of Ж, ж and к tak­ing К as a sample (with the straight double-serifed di­ag­on­al at the top and the wave-shaped bot­tom one). I re­placed the drop-shaped bot­tom ter­min­al of З with a slightly in­clined beak like the one in С and Э. I then re­placed the trapezoid­al form of Л, л with the his­tor­ic­al tri­angle form. All these cor­rec­tions made the ro­man shapes styl­ist­ic­ally more or­derly, but the Cyril­lic coun­ter­part still did not match the Lat­in mod­el ex­actly.


In the fi­nal ver­sion of the typeface, the let­ters К and Ж have the same straight up­per di­ag­on­als and un­du­lat­ing bot­tom strokes as in the sketches. The shape of the lower serifs on the З, С and Э has been uni­fied (cf. teardrop ter­min­al on the З in the above sketches).

Everything was a bit sim­pler with the it­al­ic. Un­like the ro­man out­lines, prac­tic­ally all the it­al­ic char­ac­ters were a suc­cess from the very be­gin­ning, and in later re­fine­ment all they needed was some changes in pro­por­tion, col­our, and a few de­tails. For the low case ж I chose a zig­zag struc­ture, and for ч a hand­writ­ten form from the late 18th—early 19th cen­tury. The struc­ture of some of the it­al­ic cap­it­als, however, had to be changed to match re­lated ro­man ones.

Be­ing busy with oth­er pro­jects, I had to in­ter­rupt the work on the Cyril­lic ver­sion of Kis for a while. A dis­cus­sion in 1999 with Max­im Zhukov, our con­sult­ant on Cyril­lic, helped me real­ise how to ap­proach the rest of the work. The es­sence of his com­ments was that I should stick to the his­tor­ic­al ro­man forms that cor­res­pon­ded to the gen­er­al style of the type I was cre­at­ing. I re­placed the trapezoid­al form of Д, д with a right-tri­angle struc­ture with tri­angle des­cend­ers, which is closer to the form of the 18th cen­tury. I re­placed the double-bowled struc­ture of the lower­case ф with a form close to the cor­res­pond­ing let­ter in the Civil Type, with one oval crossed by the ver­tic­al stroke, as in the Greek let­ter phi. In or­der to stick to the chosen style, the cap­it­al Ф also got the same shape as in the Pet­rine type, with an as­cend­er and a des­cend­er. To sup­port a pro­ject­ing Ф I had to draw a cap­it­al У with the tail des­cend­ing down the way it does in the Pet­rine type. And fi­nally, I re­placed the double-sided serifs at the ends of the top di­ag­on­als of lower­case ж, к with drop-shaped ter­min­als like those in the Pet­rine en­grav­ings and in the small sizes of the Civil Type.

Maxim Zhukov
Max­im Zhukov


By its nature and gen­es­is, Bit­stream Kis falls in­to a wide cat­egory of his­tor­ic­al type re­viv­als, which are very com­mon in West­ern ty­po­graphy—so com­mon that ac­cord­ing to Jerry Kelly, an Amer­ic­an ex­pert in ty­po­graph­ic design, “it may be that there are more re­vived typefaces in use now than ori­gin­al designs”. Cum grano salis, one may con­sider Pet­rovsky and Eliza­vet­in­sky typefaces of O.I. Lehmann’s foundry (1902 and 1904–07, re­spect­ively), as well as Aka­demichesky (H. Ber­thold, c. 1910), Ban­nikova Ro­man and Baikon­ur (VNIIPo­li­graph­mash, 1946–51 and 1960–69) type re­viv­als.

And yet those were pre­cisely the type re­viv­als that provided an aes­thet­ic found­a­tion for a move­ment in print design known as “ty­po­graph­ic­al re­viv­al”, ori­gin­at­ing with Wil­li­am Mor­ris. The con­tinu­ing dia­logue and end­less rivalry between two schools in West­ern ty­po­graphy of the 20th cen­tury—the tra­di­tion­al­ist “re­viv­al­ism” and the in­nov­at­ive neue Ty­po­graph­ie— were most be­ne­fi­cial for the de­vel­op­ment of book and graph­ic design. The re­viv­al­ist move­ment star­ted gain­ing mo­mentum soon after the Great War; however, Rus­sia—then already com­mun­ist—was not af­fected by it.

In ful­fil­ment of its ob­lig­a­tions un­der a con­tract with Amer­ic­an di­git­al foundry Bit­stream, Para­Type de­veloped Cyril­lic “ex­ten­sions” to a num­ber of clas­sic­al typefaces ini­tially de­signed for Lat­in script. Of all those typefaces, Kis seems to stand closest chro­no­lo­gic­ally to the first Rus­si­an “Ro­man”, the Civil Type of Peter the Great. That made it the prime can­did­ate for a cred­ible re­con­struc­tion of a Cyril­lic typeface that could have ex­is­ted in real life, as op­posed to the earli­er his­tor­ic­al typefaces—such as Aldine 401, Ori­gin­al Gara­mond or Caslon 540—whose styl­ist­ic pat­terns con­sti­tute pure and genu­ine ex­tra­pol­a­tions.

It is a well-known fact that let­ter­forms evolve over time, both in hand­writ­ing, and in print. For ex­ample, in the cap­it­al Q the long tail is pe­cu­li­ar to the Old-Style typefaces, and the short one is char­ac­ter­ist­ic of the Mod­ern styles; in Cyril­lic typefaces of the 18th cen­tury the Д and the Л were tri­an­gu­lar but by the end of the 19th cen­tury they have be­come trapezoid, al­most rect­an­gu­lar. The un­der­stand­ing of those de­tails, and giv­ing them due con­sid­er­a­tion may im­part styl­ist­ic in­teg­rity and cred­ib­il­ity to a type design.

In short, my re­com­mend­a­tions to Para­Type were to use the forms of early Rus­si­an (Pet­rine) typefaces—wherever pos­sible—as ref­er­ents for the Cyril­lic ver­sion of Bit­stream Kis.


And so, hav­ing re­tained char­ac­ters that cor­res­pon­ded to the Lat­in old style, I gave the ro­man caps some fea­tures of Cyril­lic types of the be­gin­ning of the 18th cen­tury. Sim­il­ar changes were in­tro­duced in­to the it­al­ic cap­it­als Д, Л, У, Ф. In or­der to pre­serve the styl­ist­ic unity of up­per­case Ц and Щ, I re­placed the wave-shaped des­cend­ers with small tri­an­gu­lar ele­ments sim­il­ar to those in Д. At the be­gin­ning of 2000, the work on the ba­sic Cyril­lic al­pha­bet was gen­er­ally fin­ished. Only the de­vel­op­ment of ad­di­tion­al na­tion­al char­ac­ters re­quired by stand­ard en­cod­ings (Cyril­lic, East­ern European, Turk­ish, Balt­ic) and the in­tro­duc­tion of kern­ing for com­pli­ant char­ac­ter pairs were still to be fin­ished. After the ne­ces­sary tech­nic­al pro­cessing, Para­Type re­leased Kis Cyril­lic in Post­Script Type 1 and TrueType formats for both Mac and Win­dows plat­forms at the end of Septem­ber, 2001.

  1. Original-Janson-Antiqua (1919), Stempel
  2. Original-Janson-Kursiv (1919), Stempel
  3. Janson (1932–1937), Mergenthaler Linotype, Chauncey H. Griffith
  4. Janson (1936), Lanston Monotype, Sol Hess, Bruce Rogers
  5. Ehrhardt (1938), Monotype
  6. Janson (1954), Linotype, Hermann Zapf
  7. Janson Text (1985), Linotype, Adrian Frutiger
  8. Kis-Antiqua (1980), Typoart, Hildegard Korger
  9. Janson-Antiqua, Berthold
  10. Janson SB, Scangraphic
  11. Janson Antiqua, Scangraphic
  12. Bitstream Kis (1990–1993), Bitstream
  13. URW Janson (1995), URW
  14. T-26 Turnavia (2000), T-26, Gábor Kóthay
  15. Bitstream Kis Cyrillic (2001), ParaType, Vladimir Yefimov
  16. Berthold Kis BQ (2006), Berthold
History Solo Yefimov