TATIAN'S DIATESSARON AND A PERSIAN HARMONY OF THE GOSPELS

BRUCE M. METZGER

PRINCETON THEOLOGICAL SEMINARY
for a tiny in Greek,' E. XCEPT witnesses toparchment fragmentDiatessaron all the extant Tatian's famous are of secondary or tertiary character. These witnesses may be conveniently divided into two groups, one Eastern and the other Western. The chief members of the Eastern group include, first, the Syriac commentary on the Diatessaron by St. Ephraem of the fourth century, preserved today only in an Armenian translation which has been edited from two manuscripts;' second, an Arabic Diatessaron which was translated from the
I Edited by Carl H. Kraeling, A Greek Fragmen.t of Tatian.'s Diatessaron from Dura (Studies and Documents, III; London, 1935). The editor dates the fragment about the year 222 (p. 7), that is, about fifty years after Tatian drew up the original Diatessaron. This is the only known witness to Tatian's work which is extant in Greek, for the leaf from a papyrus codex containing the Greek text of parts of Mt 18 and 19, which its editor, Otto Stegmiiller, believed'to be a fragment of the Greek Diatessaron (see his article, "Ein Bruchstiick aus dem griechischen Diatessaron (P. 16, 388)," Zeitschrift fitr die neutestamentliche Wissenschoft, XXXVII [1938], 223-229), is probably nothing more than a Greek text which contains several Tatianic readings (so Curt Peters, "Ein neues Fragment des griechischen Diatessaron?" Biblica, XXI [1940], 51-55, and "Neue Funde und Forschungen zum Diatessaron," ibid., XXIII [1942], 68-77). The Armenian text, Srboyn Ephremi matenagrouthiunk', II, was published in 1836 by the Mechitarist Fathers of the Monastery of San Lazzaro at Venice. This edition was made available for the use of scholars who are not expert in the Armenian language by J. B. Aucher who prepared a Latin rendering which was edited and published by Georg Moesinger in 1876. A collection of Ephraem's citations from the Diatessaron, arranged in the order of the Arabic Diatessaron and carefully translated into English, was supervised by J. Armitage Robinson and published as Appendix X in J. Hamlin Hill, The Earliest Life of Christ ever Compiled from the Gospels, Being the Diatessaron of Totion (Edinburgh, 1894), pp. 333-377; this Appendix, accompanied by two
JOURNAL OF BIBLICAL LITERATURE
METZGER: DIATESSARON AND HARMONY OF GOSPELS
Syriac and which is extant in two forms, represented by two and four manuscripts respectively;3 and, third, a Syriac Diatessaric lectionary for Passiontide extant in about two dozen manuscripts. 4 The chief witnesses of the Western group include, first, the famous Codex Fuldensis, a Latin harmony of the
additional essays, was reprinted with very minor alterations in J. Hamlin Hill, A Dissertation on the Gospel Commentary of S. Ephraem the Syrian (Edinburgh, 1896), pp. 75-119. According to V. F. Buchner, of the two manuscripts of Ephraem's Commentary from which the Armenian edition was prepared, it appears that manuscript A is more reliable than manuscript B; see his note, "Some Remarks on the Tradition of the Armenian Translation of Ephraem Syrus' Commentary on the Diatessaron," Bu~letin of the Bezan Club, V (1928), 34, and "Zu einer Stelle der armenischen Ubersetzung von Ephrem Syrus' Diatessaron-Kommentar," Handes Amsorya, XL (1927), cols. 685-688. 3 The editio princeps, based on two manuscripts, A of the thirteenth or fourteenth century, and B of a somewhat later date (so Paul L. Kahle, The Cairo Geniza [London, 1947], p. 213) was prepared by Agostino Ciasca (later Cardinal Ciasca), Tatiani Evangeliorum harmoniae arabice (Rome, 1888; anastatic reprint, 1930). Translations into English and German, accompanied by critical introductions and notes, were prepared by Hill (op. cit.), Hope W. Hogg, The Diatessaron of Tatian (The Ante-Nicene Fathers, IX [New York, 1896], pp. 33-138), and Erwin Preuschen with the help of August Pott, Tatians Diatessaron aus den arabischen ubersetzt (Heidelberg, 1926). The most recent edition of the Arabic text on the basis of three manuscripts (A and B with a much later one designated E) is that prepared by A.-S. Marmardji, Diatessaron de Tatien, texte arabe etab/i, traduit en fran,ais. " (Beyrouth, 1935). Unfortunately, however, it is often impossible to determine from Marmardji's apparatus whether his printed text is that of ms. E or is the editor's idea of what the ms. ought to read. For further information regarding the manuscripts of the Arabic Diatessaron, see Georg Graf, Geschichte der christlichen arabischen Literatur (Studi e testi, CXVIII; Citta del Vaticano, 1944), pp. 152-154; A. ]. B. Higgens, "The Arabic Version of Tatian's Diatessaron," Journal of Theological Studies, XLV (1944), 187-199, and Kahle, op. cit., pp. 211-228. 4 See]. P. P. Martin, Introduction it la critique textuelle du Nouveau Testament, Partie practique, III (Paris, [1885]), 121-144, and "Le.1~a. u(f(fapwv de Tatien," Rez'ue des questions historiques, XXXIII (1883), 366-378; H. H. Spoer, "Spuren eines syrischen Diatessaron," Zeitschrift fur die deutschen morgenlandischen Gesellschaft, LXI (1907), 850-859; G. A. Barton and H. H. Spoer, "Traces of the Diatessaron of Tatian in Harclean Syriac Lectionaries, " JBL, XXIV (1905), 179-195; and the appendix in Marmardji, op. cit., "EvangeIaire diatessarique syriaque," pp. 1*-75*.

Gospels prepared at the direction of Bishop Victor of Capua near the middle of the sixth century;5 second, various medieval German harmonies, the most notable of which is an Old High German (East Frankish) bilingual harmony dating from the second half of the ninth century, the Latin text of which depends upon Victor's work;6 third, the Middle Dutch (Flemish) harmonies preserved in nine manuscripts of the thirteenth to fifteenth centuries,7 the best known of which are the manuscript at LiegeS and the one at Stuttgart;9 fourth, two Old Italian
5 Edited by Ernst Ranke, Codex Fuldensis, Novum Testamentum lafine, interprete Hieronymo, ex manuscripto Victoris Capuani (Marburg, 1868). 6 Edited by Eduard Sievers, Tatian, lateinisc1t und altdelltsch mit ausfuhrlichen Glossar, 2te Auff. (Bibliothek der aUesten deutschen Litteratur-Denkmaler, V' Paderborn, 1892). For information regarding other medieval German h~rmonies, see Curt Peters, Das Diatessaron Tatians, seine Uberlieferung und sein Nachwirken im.Morgen- 1md Abendland sowie der heutige Stand seiner Erforschung (Orientalia christiana analecta, CXXIII; Rome, 1939), pp. 177-188. 7 For a list of these see Peters, Das Diatessaron Tatians, pp. 140-142. S. Edited first by G. J. Meijer, Het Leven van Jezus, een nederlandsch Handschrift uit de dertiellde Eellw (Groningen, 1835), the significance of which for New Testament scholarship was discovered fifty years later by J. A. Robinson, Academy, XLV (24 March 1894), 249-250. The manuscript was re-edited with evidence from other Middle Dutch harmonies by J. Bergsma, De Levens van Jezus in het Middelnederlandsch (Bibliotheek van middelnederlandsche Letterk1t11de, LIV, LV, LXI; Groningen, 1895-98). The lack of an index in Bergsma's volume is supplied by C. A. Phillips, Index to the Liege Diatessaron (Edition of Dr. J. Bergsma), privately printed for the members of the Bezan Club (a photostatic copy is available in the Library of Princeton Theological Seminary). It is to be hoped that the magnificent edition which has been in the course of publication under the auspices of the Royal Academy at Amsterdam will be brought to a conclusion, namely, The Liege Diatessaron, edited with a Textual Apparatus by Daniel Plooij, C. A. Phillips, and A. H. A. Bakker, Parts I-V (Verhalldelingen der koninklijke nederlandsche Akademie van Wetellschappen, Afd. Letterkunde, Nieuwe Reeks, Deel XXXI; Amsterdam, 1929-1938). For a general discussion of this and other Middle Dutch harmonies, see C. C. De Bruin, JJliddelnederlandse Vertalingen van het Niwwe Testament (Groningen, 1935), pp. 32-68, and for a stemma showing the relationship of several Dutch Harmonies, see the incisive critique of Plooij's preliminary work on the Liege Diatessaron, A Primitive Text of the Diatessaron (Leyden, 1923), by the Germanist, Th. Frings, in Literaturblatt fur gernwnische und romanische Philologie, XLVII (1926), cols. 150-155. 9 The text is printed by Bergsma, op. cit.

framework of Tatian's Diatessaron, but possess essentially a non-Tatianic form of text. In the case of the Codex Fuldensis, Victor accommodated almost perfectly the Old Latin form of text of the original to the current Vulgate. In the case of the Arabic Diatessaron, the Syriac base on which it rests is largely the Peshitta which has in most places supplanted the Old Syriac text of Tatian's harmony.I3 The chief evidence, therefore, which these two witnesses provide is not textual but structural; the frequent agreements of the sequence of sections may be presumed to reflect accurately the framework of the original Diatessaron. On the other hand, other witnesses, which are constructed according to utterly divergent sequences of Gospel material having no connection with the framework of Tatian's work, preserve Tatianic readings which were transmitted to these witnesses via the Old Syriac or Old Latin forms of text. This kind of Tatianic testimony is on a par with the type of text represented in Gospel quotations in, for example, Aphraates,'4 the Syriac Liber graduum,'5 the Armenian version and Liturgy,'6 and certain Manichaean literature'7 - all of which
13 Higgins, op. cit., shows that the form of the Arabic Diatessaron which is preserved in mss. BEO has been less thoroughly accomodated to the Peshitta than has ms. A (which latter Ciasca printed as representing the text of Tatian). q The Demonstrations of Aphraates have been edited by]. Parisot, Patrologia syriaca, I, i (Paris, 1892), ii (Paris, 1907). For Aphraates's Gospel text, see F. C. Burkitt, Evangelion da-Mepharreshe, II (Cambridge, 1904), 109-111, 180-186. IS The Syriac Liber graduum, which dates from c. A. D. 320, has been edited by M. Kmosko, Patrologia syriaca, I, iii (Paris, 1926). For the type of text in this work see A. Rucker, "Die Zitate aus dem Matthausevangelium im syrischen 'Buche der Stufen,'" Biblische Zeitschrift, XX (1932), 342-354. 16 F. C. Conybeare, "An Armenian Diatessal'on?" Journal of Theological Studies, XXV (1924), 232-245; P. Essabalian, Le Diatessaron de Tatiell et la premiere traduction des evangiles armeniens (Bibliotheque nationale, CXLII; Vienna, 1937) [in Armenian with a French resume]; and St. Lyonnet, "Vestiges d'un Diatessaron armenien," Biblica, XIX (1938), 121-150; "La premiere version armenienne des evangiles," Revue Biblique, XLVII (1938), 355-382; and "Notes philologiques sur la premiere version armenienne des evangiles," Revue des etudes indo-europeennes, I (1938), 263-270. 17 Anton Baumstark, "Ein 'Evangelium'-Zitat der manichaischen Kepha-

appear to embody in varying degrees Diatessaric readings. In fact, it is likely that the policy of approving as genuinely Tatianic only those readings in the Arabic Diatessaron which differ from the Peshitta has been unwarrantably rigorous, for even where the Arabic Diatessaron agrees with the Peshitta, if the Old Syriac also agrees, such readings are proved to be more ancient than the Peshitta and may therefore be Tatianic. Such a possibility becomes a probability with overwhelming compulsion when Ephraem and other witnesses unrelated to the Peshitta add their support. I8 To this list of witnesses to Tatian's Diatessaron another apparently must now be added, namely a medieval Persian Diatessaron of which a preliminary announcement was made several years "ago by Giuseppi Messina. I9 According to Messina this document (Laurentian manuscript XVII) was copied in the year 1547 by Ibrahim ben Shamas, a Jacobite priest, from an original dating from the thirteenth century. This earlier Persian Diatessaron appears to have been slavishly translated from a S?,riac ~ase by a Jacobite layman originally of Tabriz who calls hImself Iwannis 'Izz aI-Din, that is, "John, Glory of the Religion." Messina believes that he may have been a convert from Islam to Christianity. Although the original text of the Persian Diate.ssaron has not yet been made available, Messina has supplIed a complete table of contents and a translation into Italian of the first 71 sections out of a total of 250 (34 folios out of 130), thus comprising slightly over one fourth of the whole. It is on the basis of an examination of this material that the present article has been written.
laia," Oriens christiamts, 3. Ser., XII (1937), 169-191, and Peters, Das Diatessaron Tatians, pp. 125-132. 18 For a sane and balanced statement of the correct methodology of TatianicForse/lUng, which is drawn up with lapidary succinctness, see August Merk. Novum Testamentum graece et latine, ed. sexta (Rome, 1948), pp. 17*-18*. I, Giuseppe Messina, "Un Diatessaron persiano del seeolo XIII tradotto d~l siriaco':' Bi~lica, XX~II (1.942),268-305; XXIV (1943), 59-106, reprinted, WIth eertam mmor modIficatIOns and the addition of the translation of the remainder of the first major division of the Harmony, in Notizia su un Diatessaron persiano tradotto dal siriaco (Biblica et orientalia, X; Rome, 1943), In a subsequent study Messina deals at greater length with certain stylistic
The Persian Harmony is divided into four main divisions, containing respectively 71, 61, 60, and 58 paragraphs. The compiler has indicated the derivation of the various passages from the four Gospels 20 by using the appropriate letters, M, S (the final letter of Mar].c6s), L, and Y (Yul}.anna).2I When the sequence of the sections is compared with Tatian's work, represented in the Codex Fuldensis and the Arabic Diatessaron, only a relatively few sections are found to be in the same order, and these can be explained on the basis of natural coincidence. Indeed, the underlying plan as well as the execution seems to differ from Tatian's very carefully wrought Diatessaron. For example, the compiler of this Harmony occasionally presents parallel Synoptic passages at different places in his work (as "the salt which has lost its saltiness" Mt 513 appears in I, 34, while the parallel in Lk 1434 is given in IV, 11). At other times but one of two slightly divergent passages is utilized, the peculiarities of the other being omitted entirely in a way quite unlike Tatian's meticulous care in embodying practically everything distinctive in the four Gospels (as III, 8, where Mt 10 26b-28 is cited without the Lucan details of Lk 12 2-4). The Persian

Joseph S. Assemani, Bibliotheca orientalis, II (Rome, 1721), 159-160. Bar Salibi's statement is confirmed by evidence from Ephraem's commentary but is apparently contradicted by the Arabic text (which begins with Mark) and by the Codex Fuldensis (which begins with Luke). If the introductory notices in the Arabic manuscripts are carefully studied, however, it appears that the original Arabic text began with Jn 11. Similarly, it is almost certain that the first four verses of Luke were not in the original text of the manuscript which Victor found, for they are not mentioned in the (old) table of contents, which begins with John. '3 Theodoret, Treatise on Heresies, I, 20 (Migne, PC, LXXXIII, cols. 369-372). The two forms of the Arabic text of the Diatessaron are distinguished also (see footnote no. 21 above) by the way in which they dispose of the genealogies; in one form the genealogies are included in the midst of the text, in the other they appear at the end as a kind of appendix.
Sigla used in the Apparatus

Ar Aph E L LG Pep Per St

Arabic Diatessaron Aphraates Ephraem Liege ms. Syriac Liber graduum Pepsian Harmony Persian Harmony Stuttgart ms.
SyS SyC Syp Sypal Syhar Tus Ven Z
Sinaitic Syriac Curetonian Syriac Peshitta Palestinian Syriac Harclean Syriac Tuscan form of Italian Harmony Venetian form of Italian Harmony Zachary of Besan~on
Direct quotations from editions of Tatianic witnesses are cited in italics; translations of words and sentences into English are enclosed in quotation marks.

MATTHEW

IWCT1Jip DE
0 aV1Jp aUT1JS, DLKaLOS WV] IWCT1Jip DE aV1Jp DLKaLOs Per EVen: unde Iosep ver;ando r;a, cum ello fosse
iusto et bono L SyC -aUTW Per A I - 'Yap Per A OUDa,uws],u1J Per A L St Z (Winchester codex) OL DE aKOUCTaVTES TOU ,Ba.CTLAEws] cum audivissent (hoc) a rege Per: la parola del re A Ven SyS, c DE] IWCT1Jip Per A Ven Tus L St SyS, c, p 8ta TWV 7rPOip1JTWV] DLa TOU 7rpOip1J70U Per: per la lingua del profeta Ar L St Tus Ven: coss~ fo conpiude le profer;ie SyS, c, p, pal
-7rOTa,uw Per Tus L St E7rL XELpWV] brachiis Per: sulle loro braccia E (com) A
SyS, c, 2 mss. of p -ETL Per A SyS, c, p KEL,uEV1J] aedificata Per: fondamento sia A SyS, c, p 518 LWTa KEpaLa] Per: una parola Ven: una letera Land St: ene lettre Aph, LG, and SyS (cf. Lk 16 17): "one yod letter"

527 EPPE8TJ]+TOLS apxaLOLS

28 --r]OTJ

32 ,),aJ,tTJO'"TJ]
Per Ven Tus St Z SyC, pal, har* L SyS, C Aa{3TJ Per: prende A Ven Tus L Sys: :J~.'$ SyC, P:

29 OLEAO,),LSETO]

+ EV EaVTTJ Per: nel suo euore rifletteva (miandesid) L: wart si geturbeert in hare seluen (SyS, c hlant)

135 ')'EVVWJ,tEVOV]

itlx'4

-EO'"TWTES

Per Tus L: gaen staen LGvld. SyC, p (a hlat) 8TJO'"aVpLSETE] ponite Per: riponete A Ven Tus L Aph SyC (s hlat)

al'8E~ETaL] honorabit

Sy(s, c hlant) p, pal [ms. C] SyP (s, c hlant) 1 1

O'"OV Per a,),LOV]

+ EO'"TL
A Ven Tus L E Pep KaL Per A Ven St

ws Per Ven St

Per: onorera A SyP -ow Per A Ven Tus SyC (s hlat) 726 0J,toLw8TJO'"ETaL] OJ,tOLOS EO'"TLV Per Ven L St Z aVTWV]+KaL OL if>apLO'"aLOL Per: i loro grandi (farisei) A Tus L Z SyC, p, har (s hlat) 6

hlat) , p, har

O'"V')'')'EVELas] EV TTJ O'"V')'')'EVELa

Per A Tus L St

incipit

KaL 1I"apaXPTJJ,ta]

Per A Ven Tus L SyS (c hlat), p Per A Ven Sy' (c hlat), p, har*.

O'"WTTJPLav]

TW LEpEL] TOLS LEPEVO'"L -J,tOl'Ol'
Per A Ven Tus L E SyS, c, p, pal 2
817 l'OO'"OVS]+TJJ,tWl'
Per A Ven Tus L SyS, c, p TJVATJO'"aJ,tEV] cantavimus Per: eantammo (sarwad gujtim) A Ven Tus L SyS,C, P ow] J,taAAol' Per A SyS, C
ut liberaret nos Per: ehe ei libererebbe A SyP; ut liberavit nos Ven: salvati n'a da li nimiei Tus: aeei salvati da' nimiei nostri SyS (c hlat): "he has snatched us away unto life from the hand of our enemies" a1l"0,),paY;a0'"8aL OVO'"TJ E')'KVW E')'KVW a1l"0')'p. EKEL Per A SyS, p (c hlat)

TTJ aVTTJ] TaVTTJ EV

27 E')'El'ETO] EKTL0'"8r]
Per A L Pep SyS, p (c hlat)
Per A Ven Tus sypal Per: e lieto annunziiJ di buona speranza agli uomini A Sya (c hlat), p, har; E omits El' (2) 8] ws Per A Sya, p (c hlat) TOVTOV Per A L St SyS, p (vldentur; c hlat) aVTOl' (1)] TO 1I"aLOLOl' Per A L St Pep SyS (c hlat) p, pal LOOV Per A Tus L SyS, p, pal [maa. A, C] (c hlat) 1I"PLV TJ] EWS al' Per A SyS,P, pal (c hlat)
1I"aTTJp aVTOV KaL TJ J,tTJTTJp] TJ J,tTJTTJP aVTOV KaL IWO'"TJ<P

13 O'"OV

+ EVW1I"LOV 8EOV
Per E Aph Pep SyS (c hlat):

"for

10, God has heard the voice of thy prayer" I ')'El'l'TJO'"EL] Per: concepira e ti partorira Pep: coneeyuen& beren

25 OTL] TOVTO

Per A L Sy(s, c hlant) P: ~0'1
a')'')'EAos Per A L St Pep Sy(s, c hlant) p, Pal I 0'"0 V]

28 aVTTJl'] +0

E Aph Sy(s, c hlant) p, har
+EVAO')'TJJ,tEl'TJ O'"V El' ')'Vl'aL~LV

Per A Ven Tus L St Pep

24 See Daniel Plooij, "Traces of Syriac Origin of the Old-Latin Diatessaron," Mededeelingen der koninklijke Akaden;ie van Wetenschappen, Afd. Letterkunde, Dee! LXIII, Ser. A, No.4 (Amsterdam, 1927), pp. 20 (120) ff.

Per A: "Joseph and his mother" Tus: Gioseppo e Maria , (Ven hiat) L: Ioseph ende Maria St: Joseph ende Maria Jhesus moeder Pep : Joseph & Marie SyS, p (c hlat) 235 poJ,t<paLa] Per: lancia di dubbio E(com) !sht)'d~d of Merv (Horae Sem., V, 159) 236 STJO'"aO'"aPer: era rimasta E: "seven days she had been with a husband" (ed. Lamy, III, 813) Sys: "seven days only with a husband she was" (SyC hlat) I a1l"0 TTJS 1I"ap8El'Las] Per: vergine (bikr) St: in haren magedomme (= "in her virginity") KaL (1)] aiJTTJ Per A L St SyS, p (c hlat) 241 OL ')'OVELS aVTOV] Per: la gente di Gesu A L: Joseph ende
Maria, Pep: Joseph and Marie Tus: Gioseppo e Marie SyS (c hlat) P: "and his kinsfolk" 243 OL 'Y0VELS aVTOV] 'YJ J1-'YJT'YJp aVTOV KaL IwO''YJip Per A: "Joseph and his mother" SyP: "Joseph and his mother" 0 7r'aT'YJP O'ov Ka'Yw] eyw KaL 0 7r'aT'YJP O'ov Per A E I 00VVWJ1-EVOL1 Kat AV7r'OVJ1-EVOL Per: aifiitti con ansieta E Pep: wip myehel sorou3 Syc: "in trouble and in much perturbation" I 'YJOELTE] oLoaTE Per A Ven L St O'OipLa KaL 'YJ ALKLa] 'YJ ALKLa KaL O'OipLa Per A L: in ijaren ende in wijsheiden Z SyS, p, pal 'YVVaLKOS] +C(>LAL7r'7r'OV Per A Pep SyP. har - apxoJ1-EvOS Per A Ven Aph SyS, p (c hiat)

EJ1-(3as

OE ELS EV TWV 7r'AOLWV 0 'YJv };LJ1-WVOS. 'YJPWT'YJO'EV aVTOV a7r'O T'YJS 'Y'YJs E7r'ava'Ya'YELv OAL'YOV Ka(hO'as] KaL TO EV aVTWV 'YJV TOV };LJ1-WVOS KaL EVE(3'YJ 0 I'YJO'ovs EKaOLO'EV
Per: una nave era di Simone $aja. Gesu. sedette in queUa nave, e eomanda ehe andassero un pochino lontano daUa terra A SyS, p (c hiat) I OAL'YOV] + in aquam Per A SyS, p (c hiat) 'Yova(J'Lv] 7r'OO'LV Per A Ven Tus SyS, p (c hiat) 525 Ecp 0 KaTEKELTO] T'YJV KALV'YJV Per A Tus L Pep SyP aVTwv] aVTOV Per A Syhar mg O'OL] +J1-aO'YJTaL Per A Tus L aVTov]+WS 'YJ aAA'YJ Per Tus L St SyP, pal, har 637 KaL ov (1)] Lva Per A SyS - E'YEVETO Per A Ven L Pep SyS (c hiat) 718 TLvas Per A Ven Tus SyS, p (chiat) a'Y'YEAwv] J1-a()'YJTwv Per A SyS, p, har* (c hiat) AE'YETE]+OVTWS Per A Ven SyP -,- 7r'pOS J1-E Per A St 'YJO'YJ] 'Yap Per A Ven Tus SyS, c ipLAOV aVTOV] ipLALaV Per A Ven L mg St SyS, c, p 1112 KaL]+Eav Per A Ven Tus Sy8, c, p 1215 - opaTE Kat Per A SyS, c, p J1-EL!;ovas OLKOOOJ1-'YJO'W] OLKOOOJ1-'YJO'W KaL 7r'OL'YJO'W aVTas J1-EL!;ovas Per A Land St: sal (St: salse) meerre maken Pep: he wolde breke his berne and make it more SyS, c, p 1238 ELO'LV]+OL OOVAOL Per A Ven Tus L SyS, p

'YJv (2)] EO'TLV Per A SyC, p (s hiat) - KaL (2) Per A L St e~'YJ'Y'YJO'aTo] +'YJJ1-LV Per E L SyC, pal (8 hiat) incipit aVTOS EO'TLV 0 07r'LO'W Per A (L hiat) SyP, har I EPXOJ1-EVOS] +'YJV 'YE'YOVEV Per A Ven (L hiat) 129 (3AE7r'EL]+O Iwavv'YJs Per A Pep SyP (L hiat) (3a7r'TL!;wv] (3a7r'TL!;ELv Per: affinche battezzi A Pep Sy8, c, p - 7r'aALV Per A Ven Tus L Pep Z(Winchester codex) SyS, c, p 143 'YJ()EA'YJO'EV]+O I'YJO'ovs Per A Tus L Pep SyP - KaL (1) Per A SyS, p I ELvaL] e~eA()ELV Per E Sy6 - KELJ1-EvaL Per Tus Pep TOTE] affert Per: aUora presenta A Ven: dati [ms.: dati] Tus: e dato L: ghejt St: geijt Pep: setten jor p 211 apx'YJv] primum Per: primo A Ven: in prima L Pep SyP 327 AaJ1-(3aVELv]+aip eaVTOV Per A SyP, pal, har 332 0] KaL 0 Per A Tus L St Sy8, p, har Several of the readings in the apparatus above are worthy of more extended comments. The following remarks will serve to indicate the significance of the Persian Harmony in relation to certain Tatianic variants preserved in other witnesses. Five of the readings in the Persian Harmony reflect the embarrassment that Tatian, with his Encratite leanings,25 felt regarding certain expressions in the Gospels which refer to the relationship of Joseph to Mary and of both of them to Jesus. Thus, for example, in Mt 119 instead of representing the generally accepted Greek text, 'IWO'fIcf> 0 aVfIp aVTfjs, OLKaLOs Chv, the Persian Harmony reads e Giuseppe era un uomo giusto and thus avoids referring to Joseph as Mary's husband by omitting the Greek definite article and possessive pronoun and by taking av1}p in a general and not a marital sense. Ephraem quotes the same reading in his Commentary on the Diatessaron, "Joseph, because
25 See Daniel Plooij, "Ein enkratitische Glosse im Diatessaron; ein Beitrag zur Geschichte der Askese in der alten Kirche," Zeitschrift fiir die l1cutestamentliche Wissel1schaft, XXII (1923), 1-16 (deals with an addition to Mt 195-6).
he was a just man." Among the other medieval harmonies the yenetian Diatessaron reads unde Iosep vefando fa, cum ella fosse 'tu,sto et b.ono. It m~y be added that the Cureton ian Syriac likeWIse aVOIds offendlllg the ascetically minded and reads "Joseph because he was a just man." ". In Lk 2. there are four references to Joseph and Mary which, III the ordlllary Greek text, doubtless appeared to certain in the e~rl~ c~urch to require rephrasing in order to safeguard the :'IrgI~ bIrth.of J esus,26.In Lk 7rar~p avrou Kat, ~ p,f}rrJp IS adjusted III the Persian Harmony to read la madre di lui e Giuseppe, and in several other witnesses to Tatian the proper name "Joseph" is used in order to avoid referring to him as t , '1"\ " o 7rarrJP avrov [sc. 'IrJO'ov]. Thus, the Arabic Diatessaron reads, "Joseph a?d hi~,mother," and the Dutch Harmonies read Ioseph ende Mana (LIege ms.) and JosePh ende Maria Jhesus moeder (Stuttgart ms.). The Pepysian Harmony and the Tuscan form of the I.talian Diatessaron (the Venetian form omits this ~erse) agree wIth the Liege ms. in reading the two proper names. In Lk and 43 the phrase ol 'Y0VELS avrou is used in the Greek text in referring to Jesus' parents. In the former passage, however, the Persian Diatessaron prefers the more general term "people" in the phrase la gente di Gesit, as does also the Arabic Diatessar~n (.u~1-,)!.7 The Sinai tic Syriac (Curetonian hiat) and Peshltta lIkeWIse use a word meaning "his kinsfolk" (.C1/o!:l:ll). The Tuscan, Liege, and Pepysian Harmonies avoid the word f~; "parents" by inserting the proper names, "Joseph and Mary. In the latter passage the Persian Harmony also refuses to speak of Jesus' father and refers to la madre di lui e Giuseppe. The ~rabic Diatessaron and the Peshitta similarly abstalll from call1llg Joseph his father (but these two witnesses reverse the order, "Joseph and his mother"). In Lk the Greek text is less violently altered by the Persian Harmony. Here the words loov 0 7rarf}p O'ov Ka'Yw have resisted substitution by synonyms; only the order of words has been

26 Cf. J. Gresham Machen, The Virgin Birth of Christ (2nd ed. ; New York) ' pp. 130-131. 27 Marmardji translates, with unjustifiable laxity, "ses parents."
altered, thereby putting, significantly enough, Mary in a position of prominence (as is also the case, it will have been observed, in each of the other three passages of Lk 2 in the Persian Harmony). Both Ephraem and the Arabic Diatessaron support this inversion of order!s The Persian Harmony partially supports the famous Tatianic variant of definite ascetical import regarding Anna, the prophetess (Lk 2 36). The text, according to B~ GLXE 69 131, is rf}O'aO'a p,Era avopos ErrJ 7rra a7ro rfjs 7rap(}EvLas avrfjs. Tatlan, in accord with his Encratite tendencies, had very probably read this verse, "seven days she had been with a husband,"29 for so Ephraem refers to the passage in one of his Hymns 30 and so the Sinaitic Syriac transmits the passage (indeed, here the statement is even more emphatic by the presence of ?~, "seven days only she. "; Curetonian hiat). Though the Persian text does not reduce the conjugal life enjoyed by Anna to such a short time, it fails to render rf}O'aO'a, a word which suggests a normal married life, and transforms the married estate into a celeb ate life: "She remained seven years a virgin with her husband" (era rimasta sette anni vergine [."s::!] con suo marito). With this one may compare the Stuttgart Harmony which, instead of reading a7ro rfjs 7rap(}EvLas, has in haren magedomme ("in her virginity"),3I There were, it goes without saying, other reasons besides an ascetical tendency which prompted Tatian to make adjustments in the text of the Gospels. He was doubtless moved, for example, by literalistic considerations. When he read in Mt that Jesus "dwelt in a city called Nazareth that what had been
28 For a full discussion of these four passages see H. J. Vogels, "Die 'Eltern' Jesu (Textkritisches zu Lk. 2, 33 ff)," Biblische Zeitschrift, XI (1913),33-43. 29 Adelbert Merx argued that this reading is to be regarded as the original of this verse; Die vier kanonischen Evangelien 1tach ihrem altesten beka1tnlen Texte; II, ii, Die Evangelien des Markus und Lukas nach der syrisclzen im Sinaikloster gefmtdenen PaMmpsesthandschrift (Berlin, 1905), 207-208. 30 Edited by T. J. Lamy, Sancti Ephraem Syri ltymni et sermones, III (Mechliniae, 1889), col. 813, verse 17. 3' For a discussion of the evidence as far as it was known in 1913, see H. J. Vogels, "Lk 2,36 im Diatessaron," Biblisclte Zeitscltrift, XI (1913), 168-171. Cf. also the brief remarks by Messina, Notizia, 57-59.

spoken through the prophets (oLa rWII 7rPOCP'l/TWII) might be fulfilled, 'He shall be called a Nazarene,' " he would have been uncertain and perplexed - as others have been since his dayin attempting to discover the precise Old Testament references to which the Evangelist alludes here. So far from there being a plurality of prophets who had spoken of this matter, it is difficult enough to find in but one prophet an allusion which could have suggested to the author of the First Gospel such a predictionY Tatian, it appears, sought to avoid the multiplication of difficulties and read the singular number (oLa roD 7rpocpl]rov). This variant was perpetuated in the Tatianic tradition of the following centuries. The Arabic Diatessaron (1.5~1), the Dutch Harmonies (Liege: die prophetie; Stuttgart: den prophete), both forms of the Italian Harmonies (Tuscan: per lo profeta; Venetian: coss'i fo conpiude le profefiie), as well as the Persian Diatessaron (per la lingua del profeta) - all preserve the singular number either as "prophet" or "prophecy." The Sinai tic , Curetonian, Peshitta, and Palestinian Syriac agree in reading U:J.l!:J. Another attempt to conform a quotation to the Old Testament (in the Syriac version) appears in Mt 46. Here the promise that angels will bear one up on their hands (E7rL XLPWII) was brought into closer harmony with the Syriac Psalter (9112), which reads ~007~? ~~, "on their arms (or shoulders)"contrary to the Hebrew, C~;J~-'~' and the Septuagint, E7rL XLPWII. 33 The reading of the Persian Harmony, sulle loro braccia, is in conformity with the Arabic Diatessaron, I.5l~ ('",,~)jl, "upon their arms." Furthermore, even though Ephraem does not quote the entire promise (simply "they shall keep thee, lest at any time thy foot be dashed against a stone"), in his subsequent comments he discloses that he is aware of the tradi32 Doubtless it was the Hebrew '~l of Is 111 which supplied the Evangelist with the germinal idea developed in Mt 223; there is no evidence, however, that Tatian could read Hebrew. For a discussion of the problems involved in Mt 223, see any critical commentary, especially those by Strack and Billerbeck and by Lagrange. 33 The meaning of Xeip, "hand and arm, arm," is confined jllmost entirely to poetry and medical authors (see, in.ter alios, Stephanus, Thesaurus graecae linguae, s. v., init.).
tion regarding the arms or shoulders of the angels. 34 This variant, it may be remarked, has left no trace among the Western witnesses to Tatian, being confined to the Eastern orbit, including also the Sinaitic and Cureton ian Syriac and two manuscripts of the Peshitta. 35 In addition to variants which are due to the operation of ascetic or harmonistic tendencies, there are many others of a miscellaneous character which appear in both the Persian Harmony and in one or more of the Eastern and Western witnesses to Tatian. A Diatessaric reading which Peters detected in the West-Saxon version 36 seems to have left a trace in the Persian Harmony also. The statement in Mt regarding the Magi: 01. Of aKOVUallTS roD j3acTLAews E7rOpVO'l/uav appears in slightly divergent forms in the following Tatianic witnesses. The reading of the Venetian Harmony, e li magi aldito fio, partironssi dal re, suggests an underlying Old Latin text running something like magi cum audivissent (hoc) a rege abierunt, which reappears in the West-Saxon version tia: hi paet gebod gehyrdon, pa: fcrdon hi ("when they had heard the command, then they went"),37 Similarly in the East the Arabic Diatessaron by its reading, 1.,;;).1; 1 dUI u-- I.,. L.) ('">'oJ ("and they, when they had heard [this] from the king, went on their way"), suggests a form of the Syriac like that which is preserved in the Curetonian, ~,:. ~:o ~ 2.l?.tIo~ O~.tI ?Qo ~ ~O.l07 ("now they, when they received the command from the king, went");

pericopes (Rheinau ms. 158b, Zurich): do si des koniges rede gehorten. ; see Hans Vollmer, Verdeutsehung der Evangelien und sonstiger Teile des Neuen Testaments von den ersten AnJiingen his Luther (Bible und deutsehe Kuitur, V; Potsdam, 1935), p. 97. 38 This variant escaped Burkitt's attention in his Evangelion da-]v[epharreshe.
himself: "I will pull down my barns and build larger ones (fJ.dtovas OlKOOOfJ.~(fw)." This is represented in the Persian Harmony by distmeggero i magazzini, e edifichero e faro altri piit amplii. The Arabic Diatessaron reads, ". I will build again and will make greater ones," and the Syriac tradition (both Old Syriac and Peshitta) agrees, "I will build and enlarge (!:I~o2.) them." On the other hand, a few Tatianic readings, such as the following example, are somewhat abbreviated. In Mt the sayIng about lWTa ~v ~ KpaLa ou fJ.1J 7rapMv probably appeared in the Diatessaron in the form, "the iota (or, y6d) letter will not pass away," as is disclosed ,by the following evidence. The verse occurs twice in Aphraates and both times he quotes it, 2.?. 2.~o~2. ?o. ("one y6d letter"). This is also the reading of the Sinaitic Syriac. In the Syriac Liber graduum the saying is quoted in the same form, and the comment is added to the effect that this means the entire ten commandments, for the letter y6d is equivalent to the numeral ten. 39 In medieval times the Venetian Diatessaron read the saying in the form, una letera non se perdera. The Liege and Stuttgart manuscripts of the Dutch Harmonies read, sal ene lettre van der wet ni(e)t achter bliven. 40 In the Persian Harmony the saying appears in the form, una parola. non svanira. In Lk Tatian introduced a slight variant which unaccountably escaped von Soden's keen eye while combing the , Arabic Diatessaron for evidence of Tatian's all-pervading influence. Instead of reading "though he will not get up and give him anything because he is his friend" (. OLa TO iVaL cjJLAOV aUToO, with B~ CLX 157 etc.), Tatian substituted cjJLALav for cjJLAOV aUToO. Eastern representatives, such as the Arabic Diatessaron (:;'I.l.,all) and the Persian Harmony (a causa dell'amicizia) , agree with the testimony of Western witnesses to the Diatessaron, such as the Venetian (per l'amistade), the Liege (om de vrinschap) , and the Stuttgart (dor die vrienscap) ,

Liber graduum, xxii, 21; col. 684, ed. Kmosko. Cf. Daniel Plooij, A Primitive Text oj the Diatessaron (Leyden, 192.3), p.38.
in the support of c/>t'ALav. The Syriac tradition (Old Syriac and Peshitta) likewise preserves the word which appeared in Tatian's Diatessaron, 2.~o~. The conclusions of this study have been hinted at earlier in the article. Although this Persian Harmony discloses no relationship with Tatian's Diatessaron so far as its external framework is concerned, it is by no means worthless as a witness to the original Diatessaron. Its value for the textual criticism of the Gospels lies in the presence of many undoubted Tatianic readings which are embedded within its text. These Tatianisms show a remarkable affinity to similar readings preserved in other Eastern and Westem witnesses of the Diatessaron. As soon, therefore, as the text of the entire Persian Harmony has been made available, its evidence ought to be included in any reasonably complete critical apparatus of the Gospels.

1394 Audio Market Analysis
Digital Harmony 1394 Audio Market Analysis Confidential and Subject to NDA
Page 1 of 16 Pages 03/29/01

Introduction

For the next 2 years, Digital Harmony will focus on two primary markets for its 1394 and wireless audio technologies: Professional Audio (including sound reinforcement systems and musical instruments) and Consumer Audio (which crosses over to automotive and PC markets). For each market, a Digital Harmony 1394 Reference System has been defined. Each System contains key 1394 product categories that a) can be designed in 2001, and b) create a system to highlight the benefits of both the 1394 bus and the Digital Harmony brand. These Systems are the Digital Harmony Studio (a production system for both the professional and home recording studios) and the Digital Harmony Home (a home entertainment network). Digital Harmonys technologies are used to create interoperable networks of audio/video devices for these two primary markets. This business model is slightly different from companies that focus on creating point-to-point 1394 systems (e.g., DV camcorder-to-PC, SBP-2 peripherals-toPC, DVHS-to-DTV, DSS receiver-to-DVHS, etc.). These represent the first 1394 products to launch over the last 2 years. Digital Harmonys business model is based on the nonexclusive licensing of portfolio technologies to companies who (quite often) compete with each other, much like Dolby. To date, this nonexclusive licensing model has not been compromised. Primary sources of revenue include per-unit royalties, licensing fees (both from OEMs and semiconductor companies), and embeddable transceiver module sales (primarily for WireFree modules and DHIVA/1394 modules). Digital Harmonys expertise mirrors the background of its primary engineers and founders: highquality (as opposed to computer/Internet) digital audio. Although all of Digital Harmonys hardware and silicon design modules are engineered to support DV video and MPEG-2TS video, the companys technologies have not yet been used in a video product (e.g., DTV, DVHS). Currently, the professional audio market has embraced 1394. This market is made up of many smaller (i.e., US$10-20 million) companies, each technically sophisticated and engineering driven. 1394 audio is a relatively easy sale to this market, as there is a well-understood value/benefit to 1394. The consumer electronics market will have a slower rate of adoption, but with a much bigger market potential. It is understood that consumers are largely uninterested in the underlying technology, always in favor of benefits and value, especially broadband sharing and distributed audio/video. The initial target market (early adopters, busy affluent households) sees value in efficient use of devices, easy control, and simultaneous use by all members of the family.
Page 2 of 16 Pages 03/29/01

Digital Harmony Studio

Market: Professional Audio and Musical Instruments Main Trade Shows: AES (Sep) and NAMM (Jan) Focus: Audio Recording Products, Sound Reinforcement Products, Musical Instruments System: Project (Home) Studio, Professional Studios Leading Competitor: Yamaha (mLAN licensing program) US Market Size (1991 sales): US$3.93 billion (source: NAMM) US Market Size (1999 sales): US$6.8 billion (source: NAMM) Worldwide Market Size: approximately 2xUS market US Sales of Potential Digital Harmony/1394 Products (1999 sales): Category US 1999 Units Sold US 1999 Sales (US$) Digital Keyboards 1,298,000 $650 million Power Amplifiers 1,130,000 $583 million Loudspeakers N/A $332 million Mixers 325,000 $320 million Karaoke and DJ 1,290,000 $205 million Rack Signal Processors 508,000 $140 million Non-PC Recorders 155,000 $140 million Sound Modules 147,000 $112 million Computer Interface Products 56,000 $28 million Summary of Digital Harmony OEM Customers: Category Current Licensees

Digital Keyboards Power Amplifiers Loudspeakers Mixers Karaoke and DJ Rackmount Signal Processors Non-PC Recorders Sound Modules Computer Interface Products Peavey Crown, Crest, QSC Dynaudio, JBL, Peavey Peavey, Soundcraft, Crest Denon, Peavey Peavey, TC Electronic, dbx, Digitech, Lexicon, JBL N/A Peavey Midiman, Crest

Targeted Licensees

Roland, Yamaha, Korg, Kurzweil/Young Chang, Novation Hafler, Yamaha, Camco Genelec, Mackie, Roland, Hafler, Tannoy, EAW, Nexo Mackie, Tascam, Akai, Yamaha Roland Alesis, ART, Line6, Yamaha, Roland, Symetrix/Lucid, Presonus Korg, Yamaha, Fostex, Tascam, Roland Akai, Emu/Creative, Korg, Kurzweil/Young Chang, Roland Tascam, Event, Roland, MOTU, Sonic Foundry, Digidesign, Terratec, Echo Digital, Frontier, Aardvark, Ego Systems, Digigram
Page 3 of 16 Pages 03/29/01
Digital Audio Workstation (Computer) 8CH

Powered Monitor Speakers

2CH Legacy Adaptor (Break Out Box) 8CH
Generator (i.e. Keyboard)
Transformer (i.e. effects)
Control Surface (Asynchronous) 2CH

Control

OHCI=4 Isochronous Channels ronous
Digital Harmony Studio Reference System
Page 4 of 16 Pages 03/29/01
Important Industry Trends: 1) Active loudspeakers are replacing passive loudspeakers. Improved audio quality (i.e., mfr matches amplifier/driver to speaker), efficiency. Many mfrs such as Genelec are now making active speakers only. Sweetwater Sound quote: Were seeing a tenfold increase in sales of active speakers; the major brands selling are Mackie HR824s, Tannoy Reveals and the Event Electronics systems. Its getting very hard to sell any passive units. Sweetwater is the largest catalog/direct pro audio equipment retailers in the US. Semiconductor companies such as Cirrus Logic and TI have recently bought PWM/Class D amp technology companies. Active speakers are ideal candidate for 1394 interface (also WireFree). 2) The all-digital studio supports an arbitrary number of audio channels. Results in higher active loudspeaker sales (i.e., one per channel). Movement from stereo to 5.1, 6.1, 7.1 (Dolby, THX, and DTS), to 10.1 (Tomlinson Hulman, creator of THX). For example, Disney mixes all audio content (including music CDs) for 40 audio channels (for use in Disney theme park attractions) and then down-mixes to other commercial formats: stereo for VHS, DTS for audio CDs, Dolby Digital for DVD, ProLogic (for broadcast television). Universal operates in a similar fashion. Result: studios will have more and more monitor speakers. 3) Magnetic recording tape is being replaced with solid-state digital storage devices. Represents a crossover 1394 product category from consumer electronics (e.g., Western Digitals 61883-6 streaming HDDs) and PC (SBP-2 HDDs). 1394 HDDs are already the main choice for adding external storage to any PCbased studio, and use will grow exponentially due to greater demand for storage capacity for archiving and works-in-progress. Trend will continue indefinitely as audio formats/bandwidth grows. Already at 192kHz, MLP/DVD-Audio, multichannel (Disneys 40-channel mixes, etc.) 4) 1394 gives project studios unlimited expandability. Storage. Expanding tracking capability (going from 8 to 16 to 32 tracks when recording). Expanding mixdown format capability (adding 1394 monitor speakers to go from mixing in stereo to Dolby Digital, and so on). 5) PCI soundcards will be obsolete in the project studio environment. Inability of proprietary PCI cards to keep up with audio quality (e.g., 192kHz). Due to internal PC noise. Due to limited number of channels of I/O (mechanical layout constraints). 1394 audio interface boxes such as Digital Harmony/Crest FB-88 will dominate. 6) New 1394 product categories will emerge. 1394 (digital) microphone. (One customer has already approached DHT). Adapters that allow non-1394 devices to be connected to the 1394 system. 1394 bus power conditioners. (Monster Cable is a Digital Harmony licensee.) 7) 1394, like MIDI, will allow many innovative small companies to compete. Open standards allow small mfr to compete with the giants. 8) DJ market is surging. Seemingly retro genre is actually technology driver; non-US market in particular.

Digital Harmony 1394 Audio Market Analysis Confidential and Subject to NDA Page 5 of 16 Pages 03/29/01
Quick and Dirty Estimate of Market Potential (one approach) 1) There are an estimated 1-2 million project studios today. (Various sources.) 2) A typical project studio contains one PC-based recording system, one external multichannel audio interface, 3 rackmount signal processors, 2 digital keyboards/sound modules, one sampler module/keyboard, one audio mixer (to mix the multiple audio outputs of the sound sources), one preamplifier (for microphones and guitars), one amplifier, 2 speakers, and one backup device. 3) This represents 13 potential 1394 devices connected to a 1394-based PC workstation. 4) Each future 1394-based project studio will require at least one legacy adapter for non1394 devices. 5) Overall market today thus represents 14-28 million devices in use. 6) Each device has an expected lifetime (due to advances in technology) of 2-3 years. 7) With expected market growth rates caused by ubiquitous digital connectivity via 1394, it is estimated (Cahners and others) that there may be as many as 5 million project studios in the next 3 years. 8) With the above product life expectancy, this represents new purchase potential of 65 million 1394 devices in the next 3 years. 9) This does not take into account multiple 1394 storage devices, movement to multichannel mixes (more 1394/WireFree active monitor speakers), professional recording studios (10% of the project studio market, but using 10x as much equipment per studio), live performance systems (keyboard networks, live sound networks, architectural acoustics). 10) Of these, the commercial installation market represents a fast growing market sector, one primed for 1394. (Reference: Peavey, QSC; Competition: Peak Audios CobraNet) 11) Overall potential market is estimated to be 100 million 1394 units in the professional audio market total over the next 3 years. 12) Digital Harmony, as a world leader in 1394 audio for this marketplace, should assume high market share in terms of 1394 technologies embedded in these devices. With assumptions that only 25% of these 100 million units actually end up with 1394 ports, and a market share of 40%, a per-unit royalty of US$6/unit represents potential revenue of US$60 million over this 3-year period.
Page 6 of 16 Pages 03/29/01
Estimate of Market Potential (second approach)

TAM=Total Available Market Worldwide %1394=% of TAM that will have 1394 ports %DH=% of 1394 TAM units with DH IP Year Digital Pianos Multitrack Recorders Rackmount DSP Synths and Samplers Amplifiers Powered Speakers Interfaces Mixers Total Units 2001 TAM %1394 %DH 1,2,165 1,386 6,729 0% 15% 0% 15% 0% 15% 1% 15% 1% 25% 1% 50% 5% 50% 1% 25% 2002 TAM %1394 %DH 1,2,182 1,525 7,330 5% 15% 5% 15% 5% 15% 5% 15% 5% 25% 10% 50% 30% 50% 30% 25% 2003 TAM %1394 %DH 1,2,200 1,677 7,997 25% 15% 25% 15% 15% 15% 25% 15% 15% 25% 55% 50% 75% 50% 75% 25% 2004 TAM %1394 %DH 1,2,220 1,845 8,745 50% 15% 100% 15% 25% 20% 50% 15% 30% 30% 100% 60% 100% 60% 100% 35%
CEA, NAMM, CEDIA and internal Digital Harmony sources
SAM=Serviceable Market Worldwide=TAM*%1394 DHM=Digital Harmony Market=SAM*%DH Revenue=DH royalty revenue Year Digital Pianos Multitrack Recorders Rackmount DSP Synths and Samplers Amplifiers Powered Speakers Interfaces Mixers Total Units 2001 SAM DHM Revenue $0 $0 $0 $5 $38 $8 $25 $21 $SAM DHM Revenue $27 $8 $57 $26 $198 $101 $163 $686 $1,SAM DHM Revenue 150 1,258 2,$150 $45 $189 $144 $625 $725 $449 $1,887 $4,SAM DHM Revenue 220 1,646 $329 $198 $461 $316 $1,575 $2,056 $791 $3,874 $9,600

4,832 1,602

Notes: (1) All TAM, SAM, and DHM units are 000s (2) It is assumed that 1394 will become fairly ubiquitous in this market by 2004 (3) Digital Pianos are likely high-penetration products, since these are typically consumer, not pro, products, and can be connected to the home entertainment system via 1394 in the living room (4) Powered speakers are quickly replacing passive speakers in the pro market (10-to-1 sales increase last year), but we do not have very accurate data yet, since active speakers are so new. We are working with our speaker licensees (JBL, Peavey, Infinity, Polk, Boston Acoustics, Meridian, Linn, Revel, etc.) to get better data for the projections. (5) We expect high market share in pro audio amplifiers, since we already are working with the top 3 pro audio amp manufacturers (Crest, QSC, and Crown); we expect high market share in audio interfaces for PCs in the pro market, since we have a history of making these, and have products expected in 2001 by market leaders Peavey/Crest (largest pro audio/MI mfr in the US) and Midiman (market leader in USB audio interfaces) (6) First rackmount DSP products are expected in early 2002 from market leaders TC Electronic, Lexicon, Digitech, and dbx, all licensees. (7) A partnership with Texas Instruments or another leading 1394 semiconductor company could impact these projections significantly. Digital Harmony 1394 Audio Market Analysis Confidential and Subject to NDA Page 7 of 16 Pages 03/29/01
Rollout of Digital Harmony Studio Products: 2001 2002 2003
Four models of Crest and Midiman 1394 audio interface products for PCs Driver support in Windows, MacOS, BeOS 10Ks quantities
Continue to dominate PC interface market with new products, licensees Add signal processors (TC Electronic, dbx, others) Add amplifiers (Crest, Crown, QSC) Add architectural acoustics products (Peavey, QSC) Add active speakers with 1394 and WireFree (JBL, Dynaudio, Peavey, others) Certification testing commences (interoperability in systems) 100Ks quantities

Add remaining devices in Reference System Expand in all product categories Quantities in the millions
Wildcards: A close partnership with Yamaha could quickly create one de facto standard implementation of 1394 in this marketplace (i.e., by mating Yamahas mLAN with Digital Harmony Studio).
Page 8 of 16 Pages 03/29/01

Digital Harmony Home

Market: Audio Consumer Electronics Main Trade Shows: CEDIA (Sep) and CES (Jan) Focus: Home Entertainment Products, Whole House Audio Products System: Home Entertainment Network Leading Competitors: Zayante (Kenwood), VividLogic (Mitsubishi), SoftAcoustik US Market Size (1991 sales): US$7.63 billion (source: eBrain Market Research) US Market Size (1999 sales): US$10.28 billion (source: eBrian Market Research) Worldwide Market Size: approximately 2xUS market US Sales of Potential Digital Harmony/1394 Products (1999 sales): Category US 1999 Units Sold US 1999 Sales (US$) Direct-to-Home Satellite 3,625,000 $957 million Projection TV 1,232,000 $1,632 million DVD Players 4,072,000 $1,099 million Set-top Internet Devices 1,200,000 $192 million Rack Audio Systems 1,290,000 $205 million Separate Audio Components N/A $1,530 million Home Theater-in-a-Box N/A $229 million Home Radios 19,899,000 $348 million Aftermarket Autosound 25,872,000 $2,070 million Factory Installed Autosound N/A $2,610 million Electronic Gaming Hardware N/A $2,250 million Home Security Systems N/A $1,660 million Summary of Digital Harmony OEM Customers: Category Current Licensees
Direct-to-Home Satellite Projection TV DVD Players Set-top Internet Devices Rack Audio Systems Separate Audio Components Home Theater-in-a-Box Home Radios (also home MP3 players) Aftermarket Autosound Factory Installed Autosound Electronic Gaming Hardware Home Security Systems N/A Loewe (Sensory Science) Sensory Science, Harman/Kardon, Proceed, Meridian, Linn, Mark Levinson, Denon, Onkyo Panja, Harman/Kardon (ZapStation) Denon, Onkyo, Harman/Kardon, JBL Peavey, TC Electronic, dbx, Digitech, Lexicon, JBL, Linn, Meridian, Lexicon, Proceed, Mark Levinson Polk, Boston Acoustics, JBL, Harman/Kardon Harman/Kardon, JBL, Denon, Onkyo, Panja, iObjects, Sensory Science (SonicBlue) Denon, Infinity, JBL, Boston Acoustics Boston Acoustics, Denon, Infinity N/A Panja

Sony, Panasonic, et al. Mitsubishi, Sony Pioneer, Panasonic, Toshiba, PowerFile, et al. WebTV, ZapStation, et al. Aiwa, Sharp, Sanyo, et al. All Bose Creative, Turtle Beach, SonicBlue Pioneer, Alpine, Clarion, Kenwood Pioneer, Alpine, Clarion, Kenwood Sony, Microsoft, Sega, Nintendo Leviton, Home Director, RCS
Page 9 of 16 Pages 03/29/01
Digital Harmony Home Reference System
Page 10 of 16 Pages 03/29/01
Important Industry Trends: 1) Whole-house audio is the #1 want of Generation X/Y homeowners today, far ahead of whole-house video (most people watch movies in one location, but want to listen to music in every room of the house, the porch, the garage) 2) CEDIA is the fastest growing consumer electronics tradeshow, indicating a fast-growing market for installed home entertainment systems and home networking; key growth areas are home theater/surround sound, whole-house audio, and whole-house control. 3) Loudspeaker manufacturers embrace these trends, as they have evolved from selling each home 2 speakers (stereo), to selling them 5 speakers (DVD), to soon selling n speakers to each homeowner; in an all-digital system, there can be any number of (active digital) speakers in the house. 1394 and WireFree are ideal interfaces for this market. 4) 1394 will allow AVRs will evolve into multizone audio decoders (with 2-4 audio DSPs in each box), AM/FM/InternetRadio tuners, and legacy (non-1394 device) adapters for older sources/sinks such as phonographs, CD players, cassette decks, VCRs. 5) MP3 players need a docking station to listen to the same audio library through the home stereo system, in better-than-headphone quality. The popularity of portable MP3 players will ensure the development of the home MP3 jukebox. This will be a big new category for consumer electronics industry. 1394 is a logical connection for these, for both connecting to the AVR/decoder or 1394 speakers, and also for downloading files from gateways/PCs. 6) Convergence products (transitional products such as PVRs, Internet radios, ZapStation et al., CD burners, mobile storage gadgets, legacy adapters, video transcoders) will collectively sell well in the next 5 years before disappearing. Many of these will have 1394 so that they can also be considered PC peripherals. 7) Audio-as-product differentiator will be very prominent in devices that have 1394 for video (specifically MPEG-TS devices), as the added-value is free since the 1394 port is already present and needed for the video transport stream. Devices such as a STB or DTV with superior audio will appear; DTVs will be sold with simply add 1394 speakers directly connected to and controlled by our DTV no AVR or audio decoder needed. 8) DVD-Audio will appear and be accepted as new audio disk format. The high volumes of Universal DVD players will require custom 1394 audio/video chips to meet volumes and price points. 5C copy protection is required for audio as well. Lossless MLP audio format requires 1394 bandwidth and may require Digital Harmony audio expertise to be implemented properly. 9) Home Theater in a Box surround sound audio/speaker systems are the fastest growing segment of consumer electronics audio industry. With a 1394 port on (at least) the subwoofer/decoder, these systems can also connect to PCs. 10) Standard US retail speaker sales are flattening. The largest US speaker manufacturer and the largest Canadian speaker manufacturer are focusing on CEDIA market, and wholehouse (additional, non-living-room) speaker market with new products and marketing dollars. Extra speakers for patio and other rooms (AC power is already there) WireFree and 1394 are going to satisfy this need.

Page 11 of 16 Pages 03/29/01
Other Market Research Supporting Interconnected Consumer Audio (year 2000 sources: eBrain, CEA, others) A majority of American electronic consumers would find it useful to be able to listen to music in any room in the house from a single stereo. Four in 10 Americans of the Generation Y use their PC as much or more than a stereo to listen to music. 92% of Americans spend at least one hour a week listening to music at home, including 31% who listen for at least 11 hours per week. Among those Americans who own a stereo system, 75% report that sound quality was very important in their purchase decision. 33% of the group has connected extra speakers to the system. 60% of Americans admit to listening to music everyday. 80% of Americans claim that sound quality is very important. DVD players will continue to grow beyond their current position as the fastest selling consumer electronics product in history, with 2001 sales of 12.5 million units, up from 8.2 million in 2000. Penetration will rise to 20%. US home theater component sales will grow 16% in 2001 to US$12.5 billion, as penetration of complete home theater systems rises to near 25%. MP3 players are expected to have a 54% sales growth rate in 2001. Nearly one-third of multimedia PC owners would like to connect their regular stereo speakers to their PC. Among all college students with a PC, 3 in 10 listen to MP3 files everyday. Nearly half of the homes in America will feature a DVD player by 2006. Over half of Americans age 15-54 have played an MP3 file.
Page 12 of 16 Pages 03/29/01
Quick and Dirty Estimate of Market Potential (one approach) 1) There are an estimated 25 million US households with home theater systems today. 2) Each system has or has the capability of surround sound (5-channel) audio. 3) A typical DVD-centered home theater contains one TV, one CATV STB, one audio jukebox, one camcorder, one DVD player, one audio/video recording device, one audio/video receiver (AVR), 5 speakers, one universal remote control, and one powered subwoofer. 4) This represents 14 potential 1394 devices per home theater system. 5) Each future 1394-based home theater will require at least one legacy adapter for non1394 devices (this may become a popular feature of a 1394 AVR). 6) Overall market today thus represents 350 million devices in use. 7) Each device has an expected lifetime (due to advances in technology) of 3-5 years, which (except for the TV/display) will likely be reduced further each year as the consumer electronics industry behaves more and more like the computer industry. 8) With expected market growth rates of DVD players and surround sound audio systems, it is estimated (CEA and others) that there may be as many as 50 million US home theaters in the next 3 years. 9) With the above product life expectancy, this represents new purchase potential of 700 million 1394 devices in the next 3 years. Overall potential market is estimated to be 700 million 1394 units in the consumer electronics market total over the next 3 years. 10) This does not take into account multiple 1394 storage devices, movement to whole-house audio/video, or Internet appliances. 11) Digital Harmony, as a world leader in 1394 audio for this marketplace, should assume high market share in terms of 1394 technologies embedded in devices made by secondtier manufacturers (Harman, Denon, Onkyo, Polk, etc.) With assumptions that a) the second tier represents only 20% of the entire market (but growing), b) that 350 million units of audio devices actually end up with 1394 ports, and c) a market share of 25% of those that do, a per-unit royalty of US$6/unit represents potential revenue of US$105 million over this 3-year period.

Page 13 of 16 Pages 03/29/01
Quick Estimate of Market Potential (second approach)
TAM=Total Available Market Worldwide %1394=% of TAM that will have 1394 ports %DH=% of 1394 TAM units with DH IP Year All in One* AVR Digital TV DTV Transcoders Digital Cable STB Games DVD Players Camcorders 2001 TAM %1394 %DH 500 5,000 1,6,200 24,000 15,200 16,000 0% 10% 1% 50% 10% 25% 1% 5% 50% 5% 0% 80% 0% 1% 0% 0% 0% 1% 0% 0% 2002 TAM %1394 %DH 750 6,000 2,060 2,500 7,700 19,621 33,000 22,100 62,317 17,694 173,742 5% 50% 5% 25% 100% 15% 10% 75% 20% 5% 100% 0% 0% 40% 10% 2003 TAM %1394 %DH 1,000 6,500 3,215 6,000 9,200 25% 50% 25% 25% 100% 35% 20% 100% 100% 10% 100% 2% 2004 TAM %1394 %DH 1,400 7,000 4,340 40% 50% 45% 25% 100% 50% 40% 100% 100% 25% 100% 2% 2% 2% 2% 4% 3% 0% 100% 10%
70% 10% 10,000 2% 11,500 2% 23,200 2% 50,000 4% 35,000 3% 65,200 0% 19,500 227,140

DBS STB 17,429

0% 21,500 0% 42,000 2% 28,000 0% 63,700 0% 18,794 199,909
VCR, PVR, DVHS 61,700 Total Units 147,754
SAM=Serviceable Market Worldwide=TAM*%1394 DHM=Digital Harmony Market=SAM*%DH Revenue=DH royalty revenue Year All in One* AVR Digital TV DTV Transcoders Digital Cable STB DBS STB DVD Players VCR, PVR, DVHS 2001 SAM DHM Revenue $0 $150 $0 $8 $0 $0 $46 $SAM DHM Revenue 2,060 1,000 1,155 1,962 4,420 3,0 $113 $450 $0 $600 $0 $0 $0 $530 $0 $SAM DHM Revenue 250 1,625 3,215 4,200 3,220 4,300 42,000 6,370 18,$750 $2,438 $386 $2,520 $386 $516 $5,040 $6,720 $1,147 $SAM DHM Revenue 560 3,150 4,340 5,750 9,$1,680 $4,725 $521 $6,000 $690 $1,114 $6,000 $8,400 $2,934 $0

10,000 1,000

Games 12,000

$0 24,750

50,000 1,000 35,000 1,400 16,300 19,0

28,000 1,120

Camcorders 12,800

$0 17,694

Total Units 26,$204 56,$1,693 111,974 3,316 $19,903 153,880 5,345 $32,064 * = Home Theater in a Box complete surround sound audio system
Notes: All TAM, SAM, and DHM units are 000s It is assumed that 1394 will become fairly ubiquitous in this market by 2004 We are working with our home theater speaker licensees (JBL, Infinity, Polk, Boston Acoustics, Meridian, Linn, Revel, etc.) to get better data for the projections. We expect high market share in the higher-end AVRs due to our current plans with Denon, Onkyo, and Harman Kardon. In 2002 and beyond, these mfrs will bring Digital Harmony to lower cost, higher volume models. Harman Kardons first Digital Harmony AVR is expected to sell 100,000 units in the first year, starting some time in 2002. A partnership with Texas Instruments or another leading 1394 semiconductor company could impact these projections significantly. Page 14 of 16 Pages 03/29/01

Rollout of Digital Harmony Home Products: Assumptions: The boxes that will drive 1394 adoption are a) broadband audio/video gateways/STBs, b) digital TVs, and c) DVD players, including DVD-Audio Adoption of 1394 audio in the home entertainment system/network will proceed from the movement of portable MP3/music players from the pocket to the living room, the emergence of streaming audio in the home (Internet radios), and the resulting need for more audio decoders in more places/products. Audio jukeboxes (both hard drive-based and CD/DVD recorders and changers) will be important as well. 1394 DVD-Audio/Universal market will begin in 2002 (50 million unit forecast, as in one in every home theater) DVD-Audio music and DVD movies are only compelling with a digital audio surround sound system rendering the audio; 1394 is the likely digital interconnect, since there will already be a 1394 port on the DVD player 1394 DTVs will be available in US market in 2001 (Sony, Mitsubishi, Panasonic); three manufacturers have 1394 DTVs on the market today in Japan DTV is only compelling with a digital audio surround sound system rendering the audio; 1394 is the likely digital interconnect, since there will already be a 1394 port on the new DTVs made by Sony and Mitsubishi (market leaders in the US; Japanese DTVs from JVC, Panasonic, Sony, Hitachi, already have 1394 ports) The DVD market has exploded, already the fastest/best selling consumer electronics device of all time The DTV market is going to explode in three years, and will drag along with audio systems, as competition among DTV makers will leave them searching for additional value-added capabilities such as superior audio support The Audio-for-Video/DTV systems will likely be an A/V receiver with 1394, driving passive speakers, or a standalone audio decoder (may also be inside the DTV) with active 1394 speakers In the home, 1394 active loudspeakers will slowly take hold, as opposed to the pro audio market where they are already an established and fast-growing category of products 2001 (at risk due to availability of Crystal CS8510 chip) First 1394 AVR Complete 1394 audio/video system made by one manufacturer 1394 audio driver support in Windows, MacOS, Linux (for use in STBs, MP3 players) Reference design of 1394 home MP3/music player Reference design for 1394 audio adapter (for non-1394 audio devices) 10K quantities
2002 (full year of availability of Crystal CS8510, plus second controller chip available) AVRs from 4 licensees (3 projecting 10,000 units/month each, in first or second model) Companion DVD-Audio players from same 4 licensees Digital Harmony Composer PMA functionality built in to one STB (for system control, media library management) One standalone 1394 audio adapter 1394 MP3 jukebox for the home
Digital Harmony 1394 Audio Market Analysis Confidential and Subject to NDA Page 15 of 16 Pages 03/29/01

2003

1394 surround sound speaker system 1394 WireFree transmitter and matching speakers Certification testing commences (interoperability in systems) 100Ks quantities
Wildcards: Automotive audio (partnership with TI, market will develop quickly) PC audio (i.e., how many PC-DVD players are used at home to watch movies, creating the need for an all-in-surround sound decoder/subwoofer and satellites. A 1394 PC subwoofer/decoder with BiDAT-enabled satellites is incredibly cost effective and a compelling solution. Likely targets: Polk, Bose, Altec Lansing, JBL. Digital Harmony Composer (3 manufacturers are currently evaluating) The Composer PMA function (system control, media librarian) remains critical to demonstrating the ultimate value proposition of Digital Harmony to installers and consumers. HAVi (partnership with Philips Consumer Electronics and Philips Digital Networks)
Page 16 of 16 Pages 03/29/01