Name Recall in the Synoptic Gospels

Abstract Onomastic congruence (a feature defined in this article) is characteristic of historiographic biographies from the Early Empire. The Synoptic Gospels display onomastic congruence, as well as conservatism in their treatment of names. The preservation of names, especially those centred around key roles and events, suggests that some names may have been preserved in the oral archives of early Christian communities to footnote living eyewitness sources, paralleling historiographical situations.

systematise in the same way 'baker' or 'German' integrate into broader semantic networks. 5 Names are often meaningless.
Even in traumatic memory, when recall is most vivid and enduring, names are the earliest casualties of memory loss. 6 In ten vivid accounts of the 1944 massacre of Civitella, in which widows recollect the execution of every male from their village, only a few personal names are recalled. 7 A case where Holocaust survivors were questioned in 1984 and 1987 also revealed that many details remained fixed when comparing these interviews to those from forty years prior, while names had largely been forgotten. 8 T. M. Derico's investigations of oral traditions in northern Jordan reveal a unique exceptiona case where many personal names are recalled over a long period of time; we will return to this later. 9 But if memory is generally frail and particularly so in name recalli.e. if names are meaningless, arbitrary, and difficult to imagewhy do certain ancient works contain many authentic personal names?
This invites the further question of whether onomastics (the study of names) can be used as evidence to determine eyewitness source material in ancient texts, namely in the Gospels. Specifically, could this suggestion explain why the Gospels retain a significant quantity of appropriately distributed names corresponding at times, even orthographically, to common pronunciations? In support of this, Richard Bauckham observes that onomastics is seldom used and surprisingly little discussed in NT studies; the classicist Simon Hornblower calls it 'deplorably under-utilized' in assessing historiographical sources. 10 This paper is a small step towards remedying this defect, but it also results from initial misgivings about Bauckham's claims. First, Bauckham argues that the distribution of names in the Gospels is indicative of authenticity, because the relative frequencies of names in the Gospels reflect statistics within Tal Ilan's Lexicon of Jewish Names in Late Antiquity: Part 1: Palestine 330 BCE -200 CE (Ilan I). 11 Various elements of this invite scepticism.
First, a period of 530 years seems too broad to determine reflective patterns, and Ilan herself concedes that her lexicon does not present a 'snapshot impression of an onomastic situation'.  Moreover, could fiction in historical guise not accomplish such reflective patterns? Michael Strickland underappreciates the onomastic data of the Gospels-Acts but succeeds in demonstrating that some onomastic verisimilitude can be accomplished, for example, in the Protevangelium of James. 13 Responding to Bauckham's call to dig deeper into onomastics, but also to an impulse to investigate his claims, I compiled my own list of named individuals from the Gospels-Acts, as well as from twenty-five non-canonical works of antiquity, which were firstly comparable in length to the Gospels-Acts, and secondly centred around one or two protagonists. This created comparable onomastic data from twenty-eight sources, placing the five New Testament narratives alongside Greek romances, Greco-Roman biographies, and apocryphal materials.
My findings establish that onomastic congruence (defined below) is characteristic of historiographic biographies from the Early Empire, strengthening Bauckham's claims. Further, the Synoptic Gospels display onomastic congruence, as well as conservatism in their treatment of names.

Onomastic Congruence
'Onomastic congruence' refers to the creation of naming patterns by an ancient author that appropriately reflect the data of relevant prosopographies. The modern researcher observes it in the convergence of three factors: 1) a relatively significant number of appropriate proper names; 2) a relatively increased level of detail in proper names; 14 3) patterns of proper names reflecting 'the situation on the ground'. Seldom will a nonhistoriographical work surveyed in our study contain any one of these three features; it never contains all three. This is significant. Although it is an unconscious act by the ancient author, onomastic congruence results from a conscious historiographical impulse at some level; it is not achieved otherwise. The reasons for this are discussed below.

Relevant Prosopographies
This study refers to lexicons which, like telephone books, include lists of names but also lists of people bearing each name, often with relevant biographical details. They are onomasticons in the former sense and prosopographies in the latter, but for our interests I simply refer to them as 'prosopographies'. 15 Five are referenced: • The Lexicon of Greek Personal Names (LGPN), now published in five volumes, catalogues approximately 36,000 Greek names from 345,000 ancient people along the northern Mediterranean. 16 A sixth volume to include Palestine is forthcoming. Published volumes can be searched digitally. 13 M. Strickland, 'What's in a Name? Richard Bauckham, First-Century Palestinian Jewish Names, and the Protoevangelium of James', ATI 7 (2014) 35-42. Strickland argues that, like the Gospels-Acts, the Protevangelium of James contains first-century Jewish Palestinian names without being authentic, yet he fails to appreciate relative distribution and the qualification of popular names. 14 Typically, you need around forty names to determine patterns that can be compared with data from prosopographies, and ideally some of these names will involve qualifiers. Beyond this, several other factors impact on the ability to draw comparisons: Greek names, for example, are much more varied than Roman and especially Jewish names, meaning that you need fewer names to determine patterns for the latter. On the other hand, more detail about naming trends is often available from volumes of the LGPN which can influence the ability to draw favourable comparisons in the case of Greek names from certain provinces. 15

Patterns
Relevant prosopographies can be used to assess the authenticity of onomastic data in texts, although an appreciation for general patterns is vital for sound conclusions. David Gill's analysis of the name Δάμαρις in Acts 17.34, for example, is a good illustration of an unsound conclusion in this regard. 22 He argues that Δάμαρις is likely an invention by the author, partly because it is a singly-attested Greek name. 23 Several errors undermine this reasoning. First, the philologist Sterling Dow observes that one in twenty-five ancient Greek names is unique; Acts 16-28 contains nineteen Greek names, two of which are unique (Δάμαρις, 17.14; Λυδία 16.14), conforming to the general pattern. 24 Interestingly, Ἀπολλῶς, mentioned in Acts 18.24, is also rarely attested in the LGPN because it is not a Greek name; it is Egyptian (attested widely in ProsPtol), where Ἀπολλῶς is said to originate from according to the author of Acts. Gill's error is deepened by actual statistics now provided in LGPN 5a and 5b, which place rarity of Greek names in Coastal Asia at nine to ten per cent rather than Dow's four per cent, making Acts resemble the onomastic situation on the ground more precisely. 25 Acts 17.34 also contains another name -Διονύσιος -one of the most popular names. Διονύσιος is qualified ('member of the Areopagus') while Δάμαρις, a rare name, is not. 26 Acts 16-20 also contains three theophoric names: Τιμόθεος (16.1), Δημήτριος (19.24), and Διονύσιος (17.34). This makes three out of nineteen (15.8 per cent) theophoric, also resembling a typical onomastic situation along the ancient northern Mediterranean. 27 Although this article will focus mainly on the relative usages of common names to determine onomastic congruence, in this case Acts also demonstrates onomastic congruence through its statistically appropriate usages of rare Greek names.

Naming Practices
Naming conventions among ancient women differed from those among men. During the Roman Republic, women were generally deprived of a personal name; in ancient Athens it was against etiquette to mention a woman's name in oratory, and she was typically referred to only in relationship to a named father, husband, etc. 28 All female Latin names in Ilan I are derived from male names, with the simple addition of a female suffix (a). To give a sense of the disproportionate attention given to male names, it is worth noting that among the 2826 named persons in Ilan I (both fictional and non-fictional), only 317 are women: i.e. 11.2%. 29 Such androcentrism is pervasive, and therefore the naming practices of malesand often, of elite malesprovide the broadest data for statistical analysis, and our study interacts primarily with them only due to their statistical prominence in both the relevant narratives and prosopographies. Ancient Greek and Jewish males were typically given one name. This name was qualified most frequently by the patronym in the genitive case, where υἱός is sometimes supplied: Σώπατρος Πύρρου (Acts 20.4), Φείδωνος υἰὸς Στρεψιάδης (Aristoph. Nub. 1.134). 30 The primary name could also be qualified by deme (Ἀριστάρχου Μακεδόνος Θεσσαλονικέως, Acts 27.2), by nickname (Σίμωνος τοῦ λεπροῦ, Mark 14.3), or by other means. 31 The Roman naming system among elite males is comparatively complex. They typically had three names: the tria nomina. The first, the praenomen, was a personal name bestowed at birth; during the Republic ninety-nine per cent of males shared only seventeen praenomina. 32 Due to this shared commonality, praenomina were eclipsed in public usage by the second name: the nomen. 33 Since the nomen was the family or clan name, males were individuated by their praenomen within the household but publicly by their nomen.
Greeks struggled to assimilate this practice, however, and might refer to a public individual by the praenomen (e.g. to T. Quinctius Flamininus as 'Titus'). Due to the limitations of the praenomen, the cognomena third nameemerged as a popular alternative to qualify the nomen. 34 Since the cognomen was another personal name but far more versatile and 26 I prefer 'qualified' to 'disambiguated' since names are not necessarily qualified for the purpose of disambiguation, and for this reason, it is the least reliable criterion for assessing onomastic congruence; nevertheless, even Hornblower deems a high concentration of appropriate patronymics relevant to the discussion of eyewitness source material (Personal Names, 140). Also see the discussion below on onomastic patterns in Suetonius' Divus Julius. 27  well-suited, the nomen and cognomen together become the most common occurrence of named people, for example, in Tacitus. Praenomina were simply abbreviated or excluded in Roman literature, therefore the cognomen and nomen are the most relevant for determining onomastic patterns. 35

Onomastic Congruence in Comparative Sources
Statistics on the usage of personal names within a composition cannot be acquired through systematic computer analysis; rather, each name must be catalogued while combing through each work individually. Unintentional errors in counting are unavoidable. Michael Strickland, for example, highlights Bauckham's error in ascribing the name 'Eros' to four people in the Gospels-Acts, although Bauckham here is clearly incorrect; Bauckham did not correct this for his second edition. 36 Such a blunder, in my estimation, is not a reflection of Dr Bauckham's scholarly care (or lack thereof), and he likely remained unaware of the error; this example only serves to reflect the tedious, challenging process previously described.
The project is further complicated by other considerations: do we include nicknames? demes? patronyms? In my case, I have opted to include names that seemed standardised to the extent that they could stand alone. 37 Of course, this process involves subjectivity. Nevertheless, I am confident that the broad patterns of data that these statistics represent will not be changed by minor errors or variations in counting. Onomastic analysis of twenty-three extrabiblical compositions reveals that onomastic congruence is only found in certain biographies from the Early Empire. 38

Apocryphal Gospels
The apocryphal gospels are the least persuasive in terms of onomastic congruence. The Infancy Gospel of Thomas has eight names, with Thomas the Israelite (Θωμᾶς Ἰσραηλίτης) as the only qualified name; aside from this unnatural qualification, a Palestinian Jewish boy is named Ζήνων, a name quite common in Delos and Athens but unusual for a Palestinian Jewish child. 39 The Gospels of Peter, Mary, and the Infancy Gospel of James have no qualified names and average less than ten names per work. The Gospel of Nicodemus is the most robust with forty-six names, nine of which are qualified. However, these belong to public figures contemporaneous to the author or known from the NT. One notable exception occurs in the prologue, missing in some MSS: Joseph Caiaphas. This first name is accurate and found nowhere in the canonical literature. Other names are positively incongruous: six of twenty-two allegedly Jewish names are unattested in Ilan I (Σήμης, Δαθαής, Νεwθαλείμ, 1.1; Ἀντώνιος, Ἀστέριος, Ἀμνής, 2.4). Regarding the names of the two thieves crucified beside Jesus (Δυσμᾶς and Γέστας, 9.5); Ilan regards the former 'without plausible explanation' and the latter 'obviously a literary invention'. 40 35 Salway, 'What's in a Name?', 130. 36 Strickland, 'What's in a Name?' p. 36; cf. Bauckham, Jesus and the Eyewitnesses, p. 88 (1st ed.); in the second edition, the error occurs again in Table 6 on p. 84. 37 I would classify the name which qualifies Simon of the Gospels (Peter/Petra -'the rock') as a standardised nickname, while discounting the qualifier attached to Barsabas Justus in the Acts of Paul ('of the Broad Feet'). 38 Regarding any of the texts discussed, tables of named persons, anonymous persons, toponyms, and their intext references are available upon request. 39 This point can be overemphasised, since Jewish persons bearing this theophoric name are attested in Palestine, albeit rarely and after 70 CE (e.g. Ilan I, 281; CIIP III 2179; CIIP IV 3484). 40 Ilan I, 432-3.

Apocryphal Acts
The apocryphal Acts achieve more complex naming patterns, yet these too lack onomastic congruence. Although several names are qualified, there is no relationship between commonality of names and the presence of qualifiers. To illustrate, Barsabas Justus of the Broad Feet, Orion the Cappadocian, and Festus the Galatian -'Caesar's chief men'are qualified in the Acts of Paul (10.2); Βαρσαβᾶς, however, is a Jewish name unattested in Roman prosopography while Ὠρίων and Φῆστος, allegedly Greek, are very rare in the LGPN I-V (twelve and four attestations respectively, with zero attestations in either Cappadocia or Galatia). 41 Secondly, there is an unusually high percentage of rare Greek names; for example, the Acts of John has three rare Greek names -Κλέοβις, some MSS Κλέοβιος; Δρουσιανή; Ἀριστοβούλαout of a total of only twelve. 42 Lastly, the Acts of Peter (2.26-33) names Agrippa as prefect in Rome, while no urban prefect ever bore the name of the familiar Judean client king; 43 this Agrippa allegedly has four concubines -Ἀγριππῖνα, Νικαρία (only attested as a Greek island), Εὐwημία, and Δόριςthe latter three Greek names almost completely unattested.

Novels
Extant ancient romances, including Longus' Daphnis and Chloe, Chariton's Callirrhoe, Achilles Tatius' Clitophon and Leucippe, and Heloidorus of Emesa's Aethiopica, anticipate the modern historical novel. Daphnis and Chloe contains twenty-six names, none of which are qualified and many of which are rarely attested in the LGPN I-V; 44 further, none but two are attested in Lesbos, where the story takes place. 45 Callirrhoe has twenty-seven names. Six are qualified by deme, but there is no apparent relationship between the qualification of the name, its local setting, or its popularity; it appears to be a random feature of the text. 46 Clitophon and Leucippe contains twenty-seven names, seven of which (twenty-six per cent) are extremely rare, and it introduces several Egyptian persons with names more prominent in the LGPN I-V than in ProsPtol. 47 This situation is worsened in Aethiopica, where three of its twenty-one names are supposedly Egyptian (Ὀρουνδάτης, Aeth. 2.24; Μιτράνης, 2.24; Χαλάσιρις, 2.35), but are unattested in either ProsPtol or the LGPN I-V. Although historical novels can succeed in planting traces of verisimilitude, of 'generic markers of factuality', they fail to achieve onomastic congruence. Onomastic congruence seems to be an 'intensely 41 The context indicates that Barsabas Justus was an esteemed soldier under Nero's command; the possibility of a Jewish Roman soldier cannot be ruled out but, in any case, it is likely that the unique combination of names derives from the Barsabas Justus of Acts 1. 23. 42 The only reference to Kleobis (Cleobis) in the LGPN is to the name in Herodotus, Hist.  1 (2). Μελανθώ, 6.1.2 (4).
(even boringly) realistic' feature of a text which makes it 'difficult to sustain the classification [of fiction]'. 48 In Cyropaedia it is impossible to draw conclusions about onomastic congruence, since no two persons share a common name, and the only qualified names belong to public royal figures or military leaders. 49 In other words, there are no patterns of a 'situation on the ground'. Life of Apollonius, a composition on the verge of novel and βίος, and the Alexander Romance contain impressively complex naming patterns, although they are positively incongruent in several respects. Sometimes place names and personal names are conflated, historical figures are confused, and whole lists of names are elsewhere unattested. 50 Alex. 2.14.1 contains a list of eight members of Alexander the Great's court with names belonging to no historical persons throughout the age of Alexander; Krzysztof Nawotka comments, 'there are most probably no historical characters referred to here'. 51 Vita Apoll. 6.1-10 lists a cluster of apparently Egyptian persons, but the majority are consistently common in the LGPN and consistently rare in ProsPtol. 52

Biographies
Surprisingly, many βίοι also lack onomastic congruence. Diogenes Laertius' Life of Pythagoras, for example, does not contain enough onomastic data to be determinative. With only twenty names, neither is the Gospel of John. The Gospel of Matthew only contains onomastic congruence due to material taken over from Markan issue we discuss later. Agesilaus, like John, contains too few names (only eighteen) to be determinative.
A lack of onomastic congruence, not only in the case of the Fourth Gospel and Agesilaus, but also in the case of double tradition and M material, cannot be used to render a negative verdict on their authenticity. A lack of determinative patterns in the Fourth Gospel and Agesilaus, for example, could result from the personal nature of these works (i.e. from a lack of a reliance on named sources); further dependence on a tradition that is less narratively focused could lead to fewer names being incorporated. Certainly, the case against a composition's authenticity is increased when naming patterns are demonstrated to be positively incongruous (as with the apocryphal material discussed above), but the criterion of onomastic congruence is only relevant for the compositions that contain it. In other words, onomastic congruence positively reflects a historiographical interest (more on this below), but a lack of onomastic congruence does not disprove it.
Demonax recounts twenty-eight names and qualifies only seven; each qualified name is relatively popular and hence appropriate for qualification in a local setting, but beyond this no further patterns can be determined. 53 About half of the fifty people in Agricola have single names, while the rest are listed by their nomen and cognomen together. 54 Agricola contains onomastic congruence on two layers, although somewhat superficially; first, single names are generally rarer than qualified names, which is a natural pattern; second, three of the two common names in Agricola -Julius (4x) and Caesar (2x)are also commonly attested in the PIR; yet Nerva, attested twice in Agricola, is very uncommon. 55 Josephus' Vita has strong onomastic congruence. It names ninety-three Jewish people (109 in total), many of whom share common names: Simon (6x), Matthias (3x), Jonathan (4x), Joseph (2x), Julius (2x), Herod (4x), Agrippa (3x), John (2x), Jesus (6x), Levi (3x), Philip (2x), Ananias (2x), Justus (4x), Crispus (2x), Capellus (2x), and James (2x). Josephus qualifies all but eighteen names, most of which are comparatively rare and would need no qualification. 56 Furthermore, percentages of named people coincide well with Ilan I; a random sampling of popular names in Vita, for instance -Simon, Matthias, Jonathan, and Jesusamounts to 20.4% of named Jewish people in Vita versus 16% in Ilan I. 57 Plutarch's Caesar contains 127 named people, thirty-one qualified names and seven common names. If we focus on names in the DPRR from 110 BCE -40 BCE to create an onomastic snapshot, we discover that common names from the DPRR are typically qualified in Caesar, and that percentages of common names loosely reflect the DPRR (especially for nomina, e.g., Cornelius -3.6% of named people in Caesar versus 2.4% in the DPRR). Yet there are exceptions: Publius and Marcus, for example, amount to 1.6% and 2.4% of names in Caesar respectively, while they account for 7.7% and 11.5% of praenomina in the DPRR. 58 Suetonius' Divus Julius, like Josephus' Vita, contains strong onomastic congruence. It names 144 people, qualifies 116 names, and contains fifteen common names. The twenty-eight unqualified names are relatively rare, and distributions of common nomina and cognominaeven praenominaall generally reflect percentages in the DPRR. 59 Plutarch's most ambitious and informative biography, Pompey, also contains the most 53  extensive onomastic data and congruence from the sources surveyed. It records 172 proper names (twenty-three non-Roman), twenty-seven qualified names, and six common names, while containing several layers of reflective onomastic patterns. 60 In our survey of twenty-five sources, the only works that bear onomastic congruence are those which Craig Keener suggests mark the height of historical sensitivity for the genre of the Greco-Roman βίος, and, especially in Plutarch, the apex of this genre within the Early Empire, when expectations of historical reliability were at their highest. 61 Onomastic congruence appears to be a byproduct, however unintentional, of the information-driven nature of these historiographical works. Before revisiting this, it is helpful to appreciate that onomastic trends in the Synoptic Gospels are congruent and conservative in every accessible layer.

Onomastic Congruence in the Gospels
Onomastic trends in the Gospels are discussed by Richard Bauckham in Jesus and the Eyewitnesses. Bauckham first refines Ilan I to exclude fictional names and then compares the Gospels-Acts to the revised data. He highlights at least five examples of onomastic congruence. Firstly, popular names are generally and appropriately qualified. Secondly, Simon and Joseph are the top two male names in the Gospels-Acts and also in Ilan I. Thirdly, these two names are given to 18.2% of males in the NT narratives and to 15.6% of males in Ilan I. Fourthly, the names 'Mary' or 'Salome' apply to 38.9% of women in the Gospels-Acts; in Ilan I, the names apply to 28.6% of women. 62 Bauckham further observes that 41.5% of Jewish men in the general population, according to Ilan I, bear one of the nine most popular male names, whereas 40.3% do so in the Gospels-Acts. 63 This last observation appears quite arbitrary. Why consider the top nine names? Would choosing another arbitrary number perhaps produce a less compelling result instead? A more concerning issue is that Bauckham's maths does not seem to be correct. The top nine male names, according to his own tally of Ilan's numbers, 64 are given to a total of 1227 males (Simon, 243; Joseph, 218; Eleazar, 166; Judah, 164; Yohanan, 122; Joshua, 99; Hananiah, 82; Jonathan, 71; Mattathias, 62 = 1227). Bauckham himself notes, at the bottom of this very table, that the total number of named males counted by Ilan is 2625. But 1227 is not 41.5% of 2625; it is 46.7%. That these same nine names are held by thirty-two Jewish males in the Gospels and Acts, which amounts to 40.5% of persons (contra Bauckham's 40.3%), now seems slightly less impressive, but none of this compromises his main point. In fact, let us now take a different arbitrary number: the top six. According to Bauckham's rendering of Ilan I, 1012 males bear these six names: 38.6%. What percentage of males in the Gospels and Acts bear these same names? 34.2% (27 males), which is quite proximate.
Ilan's statistics are refined below to include only named historical Jewish males specifically datable to a range of 30 BCE -90 CE; many possible first-century names are excluded 60 Pompey contains reflective patterns similar to Caesar and Julius Divus. A sampling of qualified and unqualified names, however, further demonstrates how common names are generally qualified in Pompey while rarer names are not: the first seven qualified names, for example, are: Philippus (11 attestations in DPRR); Terentius (30); Valerius (46); Aurelius (28); Octavius (33); Calvinus (7); and Lentulus (35). The first seven unqualified names are: Cinna (8); Antistius (14); Carbo (10); Sulla (11); Vedius (1); Carrinas (3); Cloelius (4). 61  from my statistics since ostraca, ossuaries, and papyri fragments often cannot be dated narrowly, but a combination of datable names from literature, papyri fragments, and inscriptions result in a sampling of 391 male names. 65 Unlike Bauckham, I exclude persons only named in the NT narratives to ensure that the statistics are not skewed in favour of biblical texts. Our small sampling does not aim to be fully representative, meaning that an entirely independent sample to that of the Gospels is best suited for comparative purposes. The data of the top twelve names is presented in Table 1 in five columns. The first column details Ilan's comprehensive list (based on Bauckham's adjustments); the second column details all datable male names from 30 BCE -90 CE; the third details Bauckham's calculations of names in the Gospels-Acts; and the fourth and fifth columns reflect my calculations of Luke-Acts and Mark respectively (Matthew is not included because it retains many names from Mark but adds no new namessee discussion below). The comparison of Ilan's comprehensive list (column 1) to datable names from 30 BCE -90 CE (column 2) demonstrates that an onomastic snapshot of Jesus' Palestine increases several elements of onomastic congruence in the Gospels-Acts. Several names in column 1 (Manaen and Honi), which rank in the top 12 of Ilan's exhaustive list, are missing from the top names in the Gospels-Acts (column 3), but they are also missing from top datable references in Jesus' time (column 2). Alternatively, several names (Alexander and Agrippa), which are missing from Ilan's top 12 list (column 1) but attested in the top names from 30 BCE -90 CE (column 2), are also found in the Gospels-Acts (column 3).
Despite the small data sample, percentages of column 2 continue to be congruent with column 3, although certain percentages are adversely affected (esp. for Simon, Joseph, and Judas). Columns 4 and 5 demonstrate that onomastic congruence exists in Luke-Acts independently, and to a lesser extent in Mark. This is partly due to the larger amount of text in Luke-Acts, although a significant amount of the Acts' narrative involves a Hellenistic environment; nevertheless, length is certainly a factor. Regardless, this contradicts the claim that congruent-naming patterns only prevail when the Gospels are considered together. The Synoptic Gospels contain congruence independently, while John adds minor weight to the overall data. 66

Onomastic Conservatism in the Gospels
The Synoptic Gospels reveal onomastic conservatism in several respects. Bauckham notes, for example, that Matthew and Luke never add a name to an anonymous person in Mark, although they sometimes drop names. 67 This could indicate conservatism when moving from early sources (Mark) to later (Matthew, Luke). Ilan notes that NT authors follow common pronunciations rather than Hellenised orthographic practices, which leads to an almost entirely unique situation in the Gospels of the 'βαρ-' category of names 65 This number may become marginally increased with the forthcoming publication of the CIIP V, which will focus on inscriptions from Galilee. The CIIP IV includes a helpful onomastic index from inscriptions in the region of Judea, although the vast majority cannot be dated reliably within this onomastic snapshot (W. Eyewitnesses. This is not to say that John adds no weight; it contains appropriate onomastic data and adds additional unique attestations for several popular names (e.g. Lazarus, John, and Simon), but its contribution is limited by the scarcity of its data. 67 Bauckham, Eyewitnesses, 42. (Βαραββάς, Βαρνάβας, Βαρτίμαιος, Βαρσαββᾶς, Βαριωνᾶς) that reveals the Palestinian milieu of the gospels, but also that these patronyms likely served locally as nicknames. 68 Other minor patterns reveal early situational perspectives. Mark, for example, consistently places James son of Zebedee prior to his brother John (1.29; 3.16-19; 5.37; 9.2; 10.35; 13.3; 14.33). The authorial perspective clearly considers James the more prominent disciple, even qualifying John as 'the brother of James' (1.29; 3.17; 5.37) despite John's rise to prominence after the death of James around 44 CE (cf. Gal. 2.9). The relationship between Semitic style and onomastic data is also significant, although any conclusions drawn from vocabulary or syntax analysis must be held tentatively. 69 In a study of over 700 Semitisms in Luke, James Edwards concludes that Special Luke (L) material contains 400 per cent more Semitisms than materials shared with Matthew/Mark. 70 Surprisingly, this highly Semitic material also contains twenty-eight of the forty-four named individuals in Luke. This amounts to 64 per cent of named people, although L only comprises 35 per cent of that Gospel. Not only are named persons concentrated within more Semitic material, but an increase in anonymous people is evidenced in less Semitic material. 71 An independent analysis using Raymond Martin's syntax criteria produced similar results. Rather than looking for Semitisms, Martin applies seventeen syntactical criteria to determine translation Greek versus original Greek. In these criteria he considers the frequency and arrangement of certain prepositionsan approach he develops based on his analysis of over 6000 lines of Greek text from sources such as Philo, Josephus, Herodotus, Plutarch, and the LXX. 72 He produces several conclusions when he applies these criteria to the Gospels-Acts.
He concludes, for example, that Mark and Luke-Acts are combinations of translated Semitic and original Greek material. As in Edwards' analysis on Luke, Martin's criteria reveal that names cluster around the most primitive pericopes in Mark (1.1-20, 2.13-17, 3.13-19, 5.21-43, esp. 15.21-32, 16.1-8). 73 Also pertinent is Martin's analysis of Special Matthew (M), which he determines is almost entirely original Greek. 74 While he determines that M contains little to no Semitic translation material, it also contributes zero new names. 75 If one compares the onomastic data of the Gospels with Martin's analysis of original Greek versus translation Greek, this is the conclusion: a new Jewish name is never introduced in an original Greek portion of the text.
To review and summarise, onomastic trends in the Synoptic Gospels are thoroughly conservative. Onomastic congruence increases as one looks closer and earlier into these texts. Our survey of onomastic data makes it apparent that the onomastic congruence of Synoptic Gospels is comparative to the biographies of the Early Empire previously examined, and we are now prepared to consider what mechanism best explains onomastic conservatism in the Synoptic Gospels.

Explaining Name Recall
Onomastic congruence must result from a mechanism that not only conserves information but also retains apparently meaningless informationnamesin their original distributions and forms. The very nature of onomastic congruence, which appears to be an unintentional achievement of authenticity, weighs against the likelihood of a creative mechanism. Yet the irrelevance, and generally poor recall, of personal names also weighs against a strictly organic mechanism behind the phenomenon. Several considerations favour the explanation that onomastic congruence in the Synoptic Gospels results from a mechanism that retained names of living informants or guarantors of the tradition. 76 The first question facing the historian is why the ἀπομνημονεύματα of Galilean peasants should contain onomastic congruence on a par with the compositions of Suetonius, Plutarch, and Josephus. 77 Plutarch and Suetonius were known to consult archival material. Suetonius utilises senate proceedings, wills, memoirs, and the imperial libraries, while Plutarch uses fewer sourcesmemoirs, second-hand sources, eyewitness reports, letters, even oral traditionsbut does so with more discretion. 78 Josephus, too, consulted witnesses, kept notes, and wrote from a perspective of informed familiarity. 79 Onomastic congruence is reflective of their archival repository and historiographical interest. It is a plausible suggestion in the case of the Synoptic Gospels that this likewise reflects the archival repository available to them only through living tradents. It readily explains the dropping of minor characters' names from Mark's account by Matthew and Luke, the oral feature of many NT names (i.e. the Βαρ-category), and the concentration of names in primitive traditions versus the increased anonymity in later non-Semitic materials.
Again, this data can be explained by the proposal that names were significant while they referred to living informants, but were forgotten as these individuals died or became less well-known. We cannot miss a further observation: whether intentionally or unintentionally, the Evangelists were unafraid to redact layers of tradition, accumulate variations and add fresh theological colour and dialogue, but they do not add names. 80 These are treated with a unique conservatism.
Names also cluster around significant functions and events. Lists of the Twelve (Mark 3.16-19; Matt. 10.2-4; Luke 6.14-16; Acts 1.13), unlike comparable Rabbinic lists (m. 'Abot 2.8-14; b. Sanh. 43a), appear to function for the sheer purpose of conserving the names of authentic tradents. 81 While their individual significance is eventually eclipsed by their mere 'twelveness', suggesting a fading role, the lists show independence and several signs of primitivity: the prominence of James son of Zebedee over John (Mark 3.17; Matt. 10.2; Luke 6.14), Aramaisms (Βοανηργές, Βαρθολομαῖος), appropriate qualifiers, and mnemonic features. 82 The named women at the tomb are particularly significant. Carolyn Osiek argues for their primitivity, and that a generally androcentric bias rather than legal considerations (cf. Josephus, Ant. 4.219) caused the omission of their accounts until their 'explosion' into the public kerygma through the Gospels. 83 Osiek's suggestion that they are very old is confirmed by Martin's syntactical analysis of the pericopes in which these women are named. 84 The retention of specificeven variantlists of names may suggest that these women continued to function individually, as witnesses of their experiences during intervening years. 85 Some named individuals or sets of individuals were already incorporated into creedal statements shortly after Jesus' death (e.g. Peter, James, and the Twelve; cf. 1 Cor. 15.3-8). That Mark, likely also a performed text, especially retained certain names (esp. Jairus, Bartimaeus, Simon of Cyrene and his sons), suggests the possibility of their significance in early oral performances within certain communities. 86 Additionally, Kenneth Bailey and T. M. Derico observed the presence of specific informants in oral-based villages in the Middle East. As noted previously, Derico's transcripts of interviews from his ethnographical fieldwork in Jordan reveal an unusually high concentration of names. Although elements of Bailey's theory have been thoroughly critiqued by Theodore Weeden, Bailey's claim that personal names were conserved during the transmission of informal controlled tradition remains unchallenged. 87 That onomastic congruence exists in the Synoptic Gospels, with names uniquely conserved and tied to significant functions and events, suggests that some names may have been preserved specifically in the oral archives of early Christian communities to footnote living eyewitness sources, paralleling historiographical situations. 88