The Little Known Literary Battles Between the Gospel Writers (1999)
James W. Deardorff
Abstract: The most obvious solution to the Synoptic Problem (order of priority among the New Testament Gospels) has been rejected by most scholars since the 19th century largely because of embarrassments it was causing for the church and/or their own theologies. However, if theological commitment is abandoned and the evangelists are considered to have been human beings with human emotions, rather than pipelines from God, it is seen that numerous editorial oddities associated with the traditionally attested order (Matthew-Mark-Luke) are easily explainable and make good psychological sense. They indicate that the gospel writers were engaged in a behind-the-scenes tit-for-tat battle brought on by the strong anti-gentile slant of the Gospel of Matthew in its Hebraic form.
Introduction. It is unknown to most New Testament scholars, as well as to the Christian laity and clergy, that a succession of behind-the-scenes, tit-for-tat, literary battles took place as the Gospels were being written in the late 1st century or early 2nd century. At least, that is the most obvious explanation for the unflattering editorial oddities that come to light upon close comparison of the Gospels, in particular, Matthew, Mark and Luke, if they had been written in that sequence. And this is the sequence that the early church fathers placed them in, including Irenaeus, Oregin and Augustine. It is also consistent with the earliest mention of all, by Bishop Papias of Hierapolis around or shortly before mid-2nd century, that the order of certain text within Mark was not correct; this implies improper order relative to some existing written text, which then must have been that of the first attested Gospel, namely Matthew. Before we look into these literary battles, however, it would be well to first summarize why modern scholarship, since the 19th century, has preferred to assume that Mark, rather than Matthew, was the first Gospel.
The Augustinian tradition. From the beginning it was no doubt known from common knowledge, which evolved into tradition, that Matthew had come first, written in a Hebrew tongue, followed by Mark, which was an abbreviation of it, followed then by Luke, which was based upon both Mark and Matthew. This one-time common knowledge is the reason the Gospels were placed in that order within the New Testament. It could not be denied for centuries, until more modern scholarship finally decided the tradition might be all wrong. However, the reason for their radical decision seems to have been based more upon avoiding embarrassments for their own theologies and for the church than upon sound argumentation.
Problems with Mark. Since Mark is the shortest of the Gospels, it had become a worsening embarrassment, as scholarly inquiry delved ever more deeply, as to why its writer would have omitted so much from Matthew. He omitted the Sermon on the Mount (though extracting a few verses from it), and most of the parables, and he selectively omitted much Judaistic material. By postulating that Mark came first and Matthew later, however, it could be assumed that the writer of Matthew, being of a Jewish background, added this material, which could be considered to have been unavailable where Mark was written, to what he copied and altered from Mark. Then no embarrassing motivations for the editorial behavior of the writer of Mark would be relevant or arise in discussions. This reasoning was in full swing during the first half of the 19th century. It lies behind the argument that the shortest gospel must have been first.
An even greater embarrassment was thereby avoided, however. As can be seen from some nine verses within Matthew, it contains very strong anti-gentile barbs and statements to the effect that gentiles should not be, or are not worthy to be, disciples of Jesus. Perhaps two examples will suffice here:
Mt 5:47, “And if you salute only your brethren, what more are you doing than others? Do not even the Gentiles do the same?”
Mt 15:24, “He [Jesus] answered, ‘I was sent only to the lost sheep of the house of Israel.'”
Not only did the writer of Mark omit or ameliorate all of Matthew’s anti-gentile statements, but close comparison between parallel verses relating to the disciples indicates that he purposely portrayed the Jewish disciples in a derogatory light. This occurs over and over. By placing Mark ahead of Matthew in chronology, scholars could avoid this unspeakable embarrassment. It then only needed to be said that for whatever reason, the earliest gospel portrayed the disciples as being coarse, dumb or disrespectful at times, or Jesus even being disrespectful to them, and that the writer of Matthew then improved upon Mark’s language by making it more respectful and reverential. If this revision of the historical testimony of the early church fathers had not been made, as to the sequence in which the Gospels had appeared, the writer of Mark could be accused of anti-Semitism in a much stronger form than is otherwise acknowledged to exist in the Gospels.
That the New Testament reflects a dialectical tension between Jews and gentiles was brought out in 1831 by the scholar F. C. Bauer, who had argued in favor of Matthew being the first gospel, and by later scholars . However, the “tension” in mind there was between the supporters of Peter and those of Paul. The “tension” under discussion here is more blatantly expressed, but could be ignored after the Markan-priority hypothesis became dominant.
It may be mentioned that the similarities between Matthew and Mark, in the order of events in their latter sections and in wording, is so great that few scholars could hold out for their having been written independently. Thus they had come to accept plagiarism and editorial alterations as behavior commonly exhibited by the Gospel writers, and indeed plagiarism was rampant in that day. But by making Matthew dependent upon Mark, instead of vice versa, they could avoid the much greater problems outlined above.
Problems with Luke. Practically all scholars of the Gospels have had to acknowledge that Luke depends directly upon Mark. However, as long as it was thought to be dependent upon Matthew also, as had been the case up until about the 19th century due to the testimony of the early church fathers, further embarrassments for late 18th- and 19th-century scholars came to light. It had become noticed (a) that Luke most closely follows Mark’s order of events, and its wording, precisely where Mark’s order of events deviates from Matthew’s. This seemed peculiar, as why would its writer not have followed Mark most closely where it best agrees with Matthew? It was also noticed (b) that where Luke includes material present in Matthew but which is omitted in Mark, it is usually included in improper (i.e., different) context and order. Also, Luke quite often contradicts Matthew as well. Why would a writer of a Gospel, who was supposed to have behaved like a pipeline from God or in a divine manner, engage in such a disrespectful editorial manner?
To avoid these embarrassments, which mainstream scholars would like to label as non-problems that do not exist, the assumption was made, in addition to Mark coming first, that the material in Luke and Matthew not in Mark, later known as “Q”, had stemmed not from Matthew but from Q. Then they could argue that the writers of Matthew and Luke had extracted material from Q in different manners, independently of each other. By early 20th century these assumptions had become firmly entrenched, along with the additional assumption that the writers of Luke and Matthew wrote their whole gospels independently of each other. The Mark-Q priority hypothesis has remained dominant ever since, despite the lack of any literary or other evidence favoring the one-time existence of a document that could be identified as Q. Arguments were then brought forth that on the surface seemed to support Mark-Q priority, although in retrospect they are seen to be easily reversible arguments that are consistent with, or support, Matthean priority.
Thus, the theological commitment of the 19th-century scholars, plus the desire of 20th-century scholars to follow in their footsteps along the path of least theological resistance, has caused New Testament scholarship on the whole to scrap the external evidence stating that Matthew had come first and had been written in a Hebraic tongue, followed by Mark and then Luke (and then, with less certainty, the Gospel of John). Religious Studies students, for the past three-quarters of a century or more, therefore have been indoctrinated into the mode of thinking that bestows priority upon Mark and Q. Thus the scholars they grew into are scarcely even aware of the grave problems outlined here.
At the same time, the school of thought that favors Matthew-Mark-Luke order of priority faded to a very minor status. It is upheld primarily by some Catholic scholars who accept the Augustinian tradition in full, including that Matthew was written by the disciple of that name, Mark by the interpreter of Peter in Rome by that name, Luke by the physician friend of Paul, and John by the disciple of that name. For these scholars, the embarrassments discussed above are to be minimized rather than negating any part of the testimony from the early church fathers, or questioning the Gospel’s authorial attributions. Here the absence of any evidence favoring appearance of the Gospels before early 2nd century, along with the fact that they are written as if from the account of a first-hand witness though two of the three Gospels are not even attributed to disciples, is taken to mean that their authorial attributions are false. Towards the latter part of the 2nd century, however, by which time the four Gospels had received tentative acceptance, it had apparently became theologically unacceptable to claim the Gospels were not written by their attributed namesakes.
Another hypothesis, which certainly deserves mention, favored a Matthew-Luke-Mark order of gospel priority (following the work of Johann Jakob Griesbach initiated in 1783), because then the editorial behavior of the writer of Luke would not appear to be that of a crackpot, and could be explained by the writer of Mark sometimes following the order of Matthew and sometimes that of Luke, when their order of events differed. This school of thought (the Tübingen school), of which F. C. Bauer was a proponent, became a minor one, however, after the hypothesis favoring Mark-Q priority took hold. The latter hypothesis, now known as the “two-document” or “two-source” hypothesis, minimizes the most serious embarrassments for the church related to Mark and Luke while postulating and accepting as a lesser embarrassment that the Augustinian tradition was false in almost every respect.
Interestingly, the modified form of the Augustinian hypothesis favored here can be viewed as a merging of existing hypotheses. It shares with the Griesbach hypothesis the finding that certain evidence cannot be denied indicating that Matthew came first. It shares with the two-source hypothesis the finding that Luke follows Mark and is dependent upon it. It shares with the Augustinian hypothesis (and another called the Farrer hypothesis) that Luke came after Matthew and is dependent upon it. It shares with the Augustinian hypothesis that Matthew was first written in the Hebrew tongue, and with the two-source hypothesis that Mark was the first gospel written in Greek. It shares with the two-source hypothesis that the Gospels were not written by the men whose names are attached to them.
The Tit-for-tat Battles by the Evangelists
Matthew. This interesting behind-the-scenes story begins with the writer of Matthew. My own research indicates that his gospel was itself an extensive editorial revision of a source document available to him, namely the “Logia” mentioned by Bishop Papias (which modern scholars do not wish to identify as Q). Among his extensive editorial revisions, the writer of Matthew added the anti-gentile statements mentioned earlier because he was writing his gospel in the Hebraic language to convert Jews to this new Messianic form of Judaism — Christianity, and because he was apparently of the old Pharisaic mindset that detested gentiles, having once been a Pharisee and perhaps even a rabbi, judging from his wide use of the Scriptures and rabbinical writings within his gospel. He likely was active in a church within a Jewish-Christian community somewhere in the Middle East, perhaps in Antioch or Ephesus.
Mark. The writer of Mark was apparently located in Rome, according to at least two pieces of external evidence. He thus understandably favored discipleship for gentiles, which Matthew did not, so after transcriptions of Matthew reached Rome he was strongly motivated to write his own gospel. Matthew’s venue was too far away for him to have had access to the source document available to the writer of Matthew, and so he had to rely heavily upon the Gospel of Matthew as the basis for his own. He edited out or ameliorated Matthew’s anti-gentile statements and added in statements favoring gentiles. In addition, he wrote his gospel in Greek, the dominant language then of the gentile world, since his gospel was intended for gentiles.
This was the era in which Christianity was in the process of expanding rapidly into the gentile world, in the decades following the preaching of Paul, and the writer of Mark was part of this movement. He must have been appalled to read in Matthew how gentiles were being demeaned within it. (At that time, Mt 28:18-20 favoring gentiles had not yet been added; the trinitarian-like formula professed there, along with its absence from Mark and Luke, indicates it to be of relatively late date.) This situation must have been intolerable, and a gospel for gentiles written favorably for them just had to be composed. However, this writer, too, possessed very human emotions, and he could not help but allow his feelings against the writer of Matthew to creep into his text of Mark.
Mark’s tit-for-tat. In retaliation of Matthew, the writer of Mark then inserted little slurs against the intelligence and respectfulness of the Jewish disciples, thereby implying that gentiles would make better disciples than Jews. Perhaps two examples will suffice:
(1) Mk 6:48-49 states that “…And about the fourth watch of the night he came to them [to the disciples in their boat], walking on the sea. He meant to pass by them, but when they saw him walking on the sea they thought it was a ghost and cried out…”. The Gospel of Matthew, in its parallel verses (Mt 14:25-26) does not contain the clause I have italicized. Thus the Twelve are portrayed here within Mark as being unworthy for Jesus to even desire to come to their aid (this after he had set forth in their very direction, walking on the water, upon seeing that they were distressed in rowing against the wind).
(2) Mk 9:32, following the teaching by Jesus to his disciples of his coming crucifixion ordeal, states: “But they did not understand the saying, and they were afraid to ask him.” This is not present in the parallel passage of Matthew (Mt 17:23b). It indicates both mental deficiency and fear on part of the Jewish disciples, purposely added by the writer of Mark if Mark depends upon Matthew.
Full appreciation of this tit-for-tat by the writer of Mark is best realized only after studying all the numerous other similar instances. Is it any wonder, then, that a majority of biblical scholars decided to place Mark ahead of Matthew in order of priority? The absence of these aspersions in Matthew could then be attributed to that writer having avoided copying them from Mark. However, this total avoidance of what could be a harsh truth does not represent true scholarship.
This behavior on the part of the writer of Mark, if his gospel was dependent upon Matthew, seemed so distasteful that one can scarcely find anything written about it in the scholarly religious literature — it was too unthinkable to mention. Yet it could not possibly have been accidently overlooked by scholars, who revel in examining details. However, the problem of the failure of the writer of Mark to include much valuable material from Matthew, if Matthew came first, is well documented. For example, B. H. Streeter wrote that the writer of Mark would have been a “lunatic” if Matthew had been available to him and he had not included the Sermon on the Mount and most all the parables.
Luke’s tit-for-tat. The writer of Luke evidently felt the same way about Matthew’s anti-gentile barbs. However, his primary raison d’etre for bringing yet another gospel into existence was to include the valuable material from Matthew that Mark omits, thus producing a universal gospel for Jews and gentiles alike. How then could he show his intense displeasure against the writer of Matthew and his preference for Mark if he was obliged to include Matthean material omitted from Mark? Answer: by treating this (“Q”) material disrespectfully, inventing different contexts for it, and placing it any whatever order he wished.
That such editorial action by the writer of Luke had become unacceptable to most modern theologians and scholars is exemplified by the statement of Streeter that he would have to call the writer a “crank” if he had behaved editorially in this manner. Hence the desire by the dominant portion of New Testament scholarship to assume Luke to be independent of Matthew became firmly cemented into place by 1924, when Streeter wrote his book.
And how could the writer of Luke show his preference for Mark? By following Mark’s order and content most faithfully where Mark’s order of events deviates strongly from Matthew’s order (Mk 4-6). He could not, of course, have come out and expressed his feelings directly in words within a religious writing. Only subtle editorial actions were available to him. They are not so subtle, however, if one respects the testimony of the tradition that the Gospels had been written in the order Matthew, Mark and Luke, and examines the internal evidence (from redaction criticism and form criticism) and further external evidence consistent with this.
The translator of Matthew and his tit-for-tat. At some stage Hebraic Matthew was translated into Greek. It is plausible that this occurred after both Mark and Luke had been written (in Greek), but probably before mid-2nd century. This represented a golden opportunity for small alterations and additions in Matthew to be made. It had no doubt become evident to the clergy at the church, where Matthew had been written in the Hebrew tongue, that unless it were translated into Greek, its authority and popularity would give way to that of Mark and Luke, since the growth of Christianity was proceeding much more rapidly among gentiles, who generally spoke Greek, than among Jews. Although the anti-gentile writer of Matthew himself, if still alive then, probably did not feel this way, surely the more reasonable clergy did. Probably at this time, then, the translator, a different person from the original writer, added the passages of Mt 28:18-20 favorable to gentiles, probably the pro-gentile verses of Mt 12:17-22 as well, and no doubt added some reverential touches to his text. All this would make Matthew more attractive to gentiles and help counterbalance its anti-gentile statements, which could not be blatantly removed because to do such would be unfaithful to the original text. It was acceptable editorial behavior to add a bit to the text of a gospel that derived from one’s own church, but not to remove text from it. After Greek Matthew was written and transcribed, it was no doubt greatly favored over Hebraic Matthew for use within a gentile world, so the older Hebraic version and their transcriptions must have been phased out at the various churches almost as quickly as the Greek version became available.
But how then was the translator to ensure that the Greek Matthew would enjoy the authority and stature that the Hebraic version had held before Mark and Luke, written in Greek, had appeared on the scene? Mark is so much shorter than Matthew, contains so many redundant statements (pleonasms or dualisms), and contains added details (some fictional and some not, in my opinion), that it looked quite different from Matthew. Mark, having come out in Greek before Matthew did, might then seem like a fresh document, not dependent upon Matthew.
Similarly with Luke: because its material that had derived from Matthew was in different context and order, and because its writer had added so much other material of his own, it might seem like a fresh gospel independent of Matthew, and thus a gospel that would usurp the authority that Hebraic Matthew had once held, considering that Luke also preceded the appearance of Greek Matthew.
To prevent this loss of Matthean authority from occurring, the translator could do the following: he could purposely replicate strings of Greek words existing in parallel passages of Mark, on the one hand, and in parallel passages of Luke, on the other hand. Although he could not do this where distinct alterations in text from Hebraic Matthew had been made by the writers of Mark and Luke, which were frequent, elsewhere he could. This appears to be what actually occurred, and explains why strings of identical Greek words, consecutive and without any breaks, up to 33 words long are to be found within parallel passages of Matthew and Mark, and up to 27 words long in parallel passages of Matthew and Luke (most of the latter occur in the “Q” verses). By this means it would be apparent to the discerning reader that the key material within Mark and Luke was indeed dependent upon Matthew. This was the translator’s “tit-for-tat” for the anti-Matthean editorial behavior of the writers of Mark and Luke, and for their attempts to make their gospels appear as if they had not depended upon Hebraic Matthew.
That these lengthy replicated strings of Greek words did not arise by chance editorial action on the part of the translator follows from an analysis that demonstrates, for texts of the characteristic length of these gospels, that no strings of replicated words longer than about 15 are to be expected in the presence of random editorial alterations. This condition is grossly violated in the case of the synoptic gospels.
The editorial behavior on the part of the Gospel writers revealed here may seem distasteful to Christians, but is not much worse than what can be found even today among writers of opposing views, e.g., politicians of opposing parties. And given that the less-than-ethical practice of plagiarism was common 1900 years ago, it should not seem so surprising that a behind-the-scenes battle as outlined here actually took place.
Besides what has been mentioned here, there are other embarrassments involved with the origins of the Gospels. Among these is the late date at which any of the Gospels is first mentioned by name (around A.D. 130 or 140) or even referred to through direct quotation of verses, the probable fact that they were not written by the men whose names are attached to them, the great dearth of written literature that would explain anything about who wrote them, when, where and why, and why the source material available to the writer of Matthew never survived nor the five treatises written about the Logia by Bishop Papias. However, this is another story, and the little known battle between the evangelists is embarrassment enough to report here.
 E.g., see W. Barnes Tatum, The Quest of Jesus, A Guidebook (Atlanta, GA: John Knox Press, 1982) pp. 14-16.
 Papias as referenced by Eusebius, Ecclesiastical History (EH), 3.39.15; or see Tatum, Quest for Jesus, pp. 12-13. Papias lived in the first half of the 2nd century; Hierapolis was located in present-day Turkey, east of Ephesus. Eusebius was the famous historian and bishop of Caesarea who lived around the end of the 3rd century.
 Which Gospel had come first, second and third, could hardly not have been known at the time. At the church where Matthew was written, those in on its writing would have known Matthew was the first orthodox gospel, written in the Hebrew tongue; so also would those at other churches to which transcriptions of Matthew were disseminated before Mark came out. At the church in Rome where Mark was written those aware of its writing would certainly have known that Matthew preceded Mark. In the church where Luke was composed those aware of its writing would have known that Matthew and Luke preceded it, as would similarly the other churches possessing Matthew and Mark to which transcriptions of Luke were sent. Since this Matthew-Mark-Luke sequence had been common knowledge and was not in itself heretical, it grew into tradition which the early church fathers would have had no reason to falsify. Only many centuries later, after scholars were well versed in comparing parallel passages between gospels, might scholars get away with assuming that the tradition had been in error.
 Actually, the fact that Mark is the shortest gospel, such that its contents are mostly contained within Matthew and Luke, was a primary argument put forth in the initiation of the Mark-first hypothesis. E.g., see Hado Uden Meijboom, A History of the Origin of the Marcan Hypothesis 1835-1866 (University of Groningen, The Netherlands: 1866; available from Macon, GA: Mercer University Press, 1993, J. Kiwiet, transl.) pp. 14-49. Influential scholars Eduard Ruess and Heinrich Ewald were among the most theologically committed in that era.
 These are Mt 5:47, 6:7, 6:32, 10:5, 10:18, 15:24, 15:26, 18:17 and 24:9b.
 Quoted from the RSV Bible (1952).
 See Meijboom, Origin of the Marcan Hypothesis, p. xxii; or F. C. Bauer, “Die Christuspartei in der korinthishcen Gemeinde,” Tübingen Zeitschrift für Theologie 5 (1931) pp. 61-206.
 E.g., see Donald Guthrie, New Testament Introduction, 3rd Ed. (Downers Grove, IL: Inter-Varsity Press, 1970) p. 94.
 Joseph Zias, response to reader in Biblical Archaeology Review 25 (March/April 1999) pp. 10, 16. Zias had Galen, the Greek physician of the 2nd century, and Pliny the Elder in mind here, among others.
 This stems from the fact that the order and content of Lk 8:4-9:17 follows that of Mk 4:1-6:43 so well, and again from Lk 9:18-9:50 relative to Mk 8:27-9:40. Since it had already been assumed that Mark came before Luke, the dependence between them had to be that of Luke depending upon Mark.
 It is in this same region of Mark that its order deviates greatly from that of Matthew. The phenomenon from the viewpoint of the two-source hypothesis is “Wherever Matthew departs from Mark’s order Luke supports Mark, and wherever Luke departs from Mark, Matthew agrees with Mark,” as stated by Burnett H. Streeter, The Four Gospels (London, 1924; London: Macmillan Co., 1964, p. 161). From the viewpoint of the Augustinian hypothesis, the same phenomenon is stated: Wherever Mark departs from Matthew’s order Luke supports Mark, and wherever Luke departs from Mark, Mark agrees with Matthew.
 This includes the following pericopes: Sermon on the Plain/Mount, The Lord’s Prayer, God Answers Prayer, The Eye is the Lamp, Sign of Jonah, Woes Against the Pharisees and Lawyers, Anxiety Against Earthly Things, Jesus Came Not to Bring Peace, Interpreting Signs of the Weather, Settling with One’s Accuser, Cutting Down the Fig Tree, Enter by the Narrow Door, A Prophet Must Die in Jerusalem, Humility for Guests, Parable of the Great Supper, The Way of the Cross (2nd time), Parable of the Lost Sheep, True Stewardship, The Law and John, On Divorce, and Parable of the Pounds/Talents.
 Eusebius, The Theophania (Cambridge: University Press, 1843) p. 14. Here Eusebius quotes from a Gospel and refers to the passage as “the glorious words of the divine men (so) teaching.”
 However this assumption does run into the problem that Q is defined to be the material shared by Matthew and Luke absent from Mark, but the writer of Mark is considered to have been aware of some of it (e.g., the Wilderness Temptation Story, present in Mark in only a couple of summarizing sentences).
 H. G. Jameson, The Origin of the Synoptic Gospels (Oxford: Basil Blackwell, 1922); B. C. Butler, “The Synoptic Problem” in A New Catholic Commentary on Holy Scripture, R. C. Fuller, L. Johnston and C. Kearns, et al., eds., (Nashville, TN: Nelson, 1969) pp. 815-821; James Deardorff, The Problems of New Testament Gospel Origins (Lewiston, NY: Edwin Mellen Press, 1992) pp. 121-156.
 Arthur J. Bellinzoni, “The Gospel of Matthew in the second century,” The Second Century (J. of Early Christian Studies) 9 (No. 4) (1992) pp. 201-217.
 Bellinzoni, “Gospel of Matthew in the second century,” pp. 217-254.
 Papias as reported tersely by Eusebius, EH, 3.39.16. The translated statement is: “Matthew compiled the Logia in the Hebrew language, and each interpreted them as best he could.” This one terse sentence is all that has been reported or has survived about the first Gospel’s origin, which really ought to have received much literary write-up due to the gospel’s great value to the church.
 That Peter and (John) Mark in Rome had an important written document with them was attested by Clement of Alexandria (Eusebius, EH 6.14.6-8) and Justin Martyr (Dialogue with Trypho 106.9-10).
 Pierson Parker, “The posteriority of Mark,” in New Synoptic Studies, W. R. Farmer, ed. (Macon, GA: Mercer University Press, 1983) pp. 80-87; or see Deardorff’s website article at: <URL:http://www.proaxis.com/~deardorj/mksecond.htm#Mk6>.
 Streeter, The Four Gospels, p. 158.
 Streeter, The Four Gospels, p. 183.
 See Deardorff, Problems of New Testament Gospel Origins.
 Just in the Markan text paralleled by Matthew, Mark contains 79 pleonasms not present in Matthew, by one count. See C. M. Tuckett, The Revival of the Griesbach Hypothesis (New York: Cambridge University Press, 1983) p. 20. These “changes for the sake of change” were apparently fed in by the writer of Mark so that his text would look different from that of Matthew.
 Deardorff, Problems of New Testament Gospel Origins, pp. 52-57.
 See Deardorff’s website file at: <URL:http://www.proaxis.com/~deardorj/priority.htm#II>.