¹ The first systematic and extensive attempt to sketch out the career of Hennas was made by Theodor Zahn, Der Hirt des Hennas, Gotha, 1868. It performed two fundamentally important services: (1) it insisted on the matter-of-fact reality of the many personal incidents and references, whereas previous critics had been inclined to regard the whole work as an allegorical fiction; and (2) it discredited the testimony of the Muratorian Canon, which explicitly dates the Shepherd in the days of Pius, bishop of Rome, about 140 A.D. With the internal chronology of the book Zahn was less successful, even going so far as to put Sim. viii ahead of Vis. iii in point of time (p. 193). In the general series of Patrum Apostolicorum Opera by Gebhardt, Harnack, and Zahn, the volume on Hermae Pastor, Leipzig, 1877, offered the Greek text as established by Gebhardt and a commentary by Harnack. It was long the best edition, but in so far as the biographical data were concerned it moved backward rather than forward from the position of Zahn. The first Vision, for example, Harnack regarded as more or less fictitious, and he accepted the late date. Vis. ii. 4, 3 contains a reference to Clement, apparently the author of the Roman epistle to Corinth written about 95 A.D.; but this point Harnack was disposed to explain away. Later, in his Geschichte der altchrist-lichen Litteratur, II. 1, p. 265, he admitted the connection as plausible. G. Salmon, art. ‘Hermas,’ Dictionary of Christian Biography, London, 1884, II, pp. 912–921, argued ably for the early date on the strength of the mention of Clement, and maintained that the book was a collection of actual dreams or of dream-like visions, dressed up somewhat for presentation in writing. See the same author's Historical Introduction to the Books of the New Testament, 7th ed., London, 1894, pp. 527–551. This sound and reasonable view of the Shepherd was much furthered by H. Weinel, a critic deeply versed in the modern science of psychology, who wrote the treatment of Hermas in E. Hennecke's Neutestamentliche Apokryphen, Tubingen, 1904, pp. 199–210, 217 ff., and the commentary in Hennecke's Handbuch zu den neutestamentl. Apokryphen, 1904, pp. 290 ff. The most brilliant characterization of Hermas that has appeared in any language is by C. Bigg, Origins of Christianity, Oxford, 1909, pp. 72–84. And yet, though essentially sound, it is somewhat hard and disdainful, lacking the sympathetic and appreciative touch of Weinel, whose commentary, all things considered, must be pronounced the most valuable since Harnack's. A useful summary and judgment of previous opinions is given by F. X. Funk, Patres Apostolici, 2nd ed., Tubingen, 1901. The most convenient edition for ready reference is by K. Lake in the Loeb Classical Library, 1913, where the text and a new translation of the Shepherd are printed on confronting pages.

² “The man who brought me up,” says Hermas in his introduction, “sold me to a certain Rhoda at Rome. Many years later I met her again (ἀνɛγνωρισάμην) and began to love her as a sister.“The passage has been the cause of much discussion among critics. Some have even taken it for an imitation of the current type of Greek love-story: the author comes to Rome, a recognition occurs, and love follows. It is true that the recognition-motive (άναγνωρισμός) was a stock element in the romances of the New Comedy, as represented by Menander and imitated by Plautus and Terence; but the love to which Hermas refers was of a different sort. The phrase “began to love her as a sister” means simply that he and Rhoda met again within the Christian community, in which all were regarded as loving members of a single family. The common meals were called ἀγάπαι; ἀδɛλϕός and ἀδɛλϕή) were stated terms for the Christian man woman respectively. Note especially 1 Cor. 9, 5: “Have we not the right to lead about a sister [i.e., a Christian woman] as wife?” See also Torrey's comments on the appositive ἀδɛλϕί in Acts 15, 23 (The Composition and Date of Acts, 1916, p. 89). But Vis. ii. 2, 3; 3, 1, show that in the mind of Hermas there was an ascetic implication in the term, and the symbolism of Sim. ix. 11, 3 reflects the peculiar and at times doubtless perilous tendency which this lofty spiritual affection might take.

From a biographical point of view the introduction makes a strange choice of facts, and the omissions seem still stranger – the failure to mention the death of Rhoda or the escape of Hermas from slavery. Certain critics therefore conclude that the original introduction has been lost. But Hermas was not constructing an orderly “vita,” but preparing to explain a mystical vision; and he narrates only those things which serve his purpose. He proceeds, not indeed with literary art but with a certain deadly directness, toward the main point, giving the merest outline of facts regarding his acquaintance with Rhoda. Thus understood, his introduction leaves little to be desired, and it is not probable that anything has been lost.

³ Vis. i. 1, 7. Codex Sinaiticus has θεάν supported by the older Latin version. Codex Athous, which shows a consistent tendency to make Hennas appear less pagan toward the beginning and more orthodox toward the end than he probably was, reads θυγατέρα. ‘daughter.’ The Aethiopic has ‘my mistress,’ standing presumably for κυρίαν. Hilgenfeld (1881) suggests an improvement that the ancient scribes never thought of, namely, θείαν. ‘eine Tante.’ The references to Cumae and the Sibyl caused the copyists a similar uneasiness.

⁴ The date of Hermas is commonly discussed as if it were a single point in time, whereas nothing can be plainer than that the book grew by a process of accretion during a number of years. On the strength of the reference to Clement in Vis. ii. 4, 3, Zahn, Salmon, and Bigg have argued for a date about 95 A.D., and have refused to accept the external testimony of the Muratorian Fragment – a tradition that seems to be repeated with slight variations in the Liberian Catalogue of the bishops of Rome and in the anonymous Carmina adv. Marcionem iii. 294 f. The date about 140 is no longer commonly defended; see, for example, B. H. Streeter, The Four Gospels, London, 1924, p. 528: “Personally, in view of the arguments of Salmon and Bigg I incline to date Hermas c. 100, and to regard the statement in the Muratorian Canon as a by-product of anti-Montanist polemic.” Nevertheless, the references to persecution, especially in Sim. ix. 28, reflect just such a situation as prevailed under Trajan – Christians haled into court and tried for their faith as such. Hence the authors of the three most important commentaries on Hermas – Harnack, Funk, and Weinel – hold that he wrote during or after the time of Trajan, probably after 111 or 112, when Pliny in Bi-thynia asked the emperor's instructions as to the treatment of Christians. But the deadlock is not hopeless; both parties are probably right. It is the Visions that contain the reference to Clement; these therefore belong almost certainly under Domitian. It is the Similitudes, particularly viii and ix, that reflect such conditions as are set forth in the letters of Pliny and of Ignatius; and these therefore may be dated, though with somewhat less certainty, under Trajan.

⁵ Of especial interest are the very pointed allusions to ecclesiastical strife in Vis. iii. 9, 2. 7–10. The selfish ambitions of the leaders of the church are scathingly rebuked. If directed against the Roman presbyters, the passage lacks tact, to say the least. But if Vis. iii is really contemporaneous with 1 Clement, the warnings take on another significance. Hermas probably heard the letter to the Corinthians read before the Roman congregation, was impressed with the gravity of the schism in Corinth, and felt moved to do his part also toward correcting the situation. It was, indeed, just at this time that his ambition became world-wide; Vis. ii. 4, 2 sets forth a plan to make known these revelations to “all the elect.”

⁶ Some critics, it is true, have seen here a reflection of actual Arcadian scenery. “Mr. Mahaffy reports (Rambles in Greece, p. 330, 2nd ed.) that the scenery he [Hernias] describes is such as may be witnessed in Arcadia, and is not to be seen in the neighbourhood of Rome” (Salmon, ‘Hennas,’ Diet, of Christian Biography, London, 1884, p. 920). J. Rendel Harris, ‘Hermas in Arcadia,’ Journal of Biblical Literature, June 1887, pp. 69–83, makes out a remarkable case for at least two of the twelve mountains, but recognizes that many, if not most, of the others are artistic creations of Her-mas for didactic-allegorical purposes. Harris finds the language of Hermas in close correspondence with that of Fausanias's description of the region around Orchomenos. Though the case is very skilfully presented, the connections with Pausanias seem more than dubious. The dates entirely forbid the supposition that Hermas had read the Itinerary. But he may have been in Arcadia at some time. It is a shrewd guess of J. Armitage Robinson, Collation of the Athos Codex of the Shepherd of Hermas, 1888, pp. 30 ff., that Hermas was brought up in that vicinity. This is a much better explanation of the visit to Arcadia than the theory of Reitzenstein, Poimandres, p. 33, that it is modelled on some Hermetic document now lost. D. Völter, Die apostolischen Väter, Leyden, 1904, thinks the Arcadian scenery better suits the plain of Pheneos than that of Orchomenos.

⁷ The suggestion that the Visions were written by the apostolic Hermas of Rom. 16, 14, and the rest of the book by the brother of Pius, seems to have been made independently by Thiersch, Die Kirche im apostolischen Zeitalter, Frankfurt and Erlangen, 1858, pp. 352 ff., and by de Champagny, Les Antonins, 2nd ed., Paris, 1863, I, p. 144, followed by D. Guéranger, St. Cécile et la Société Romaine aux deux premiers siècles, 2nd ed., 1874, pp. 132 ff., 197 ff. Haussleiter, ‘De versionibus Pastoris Hermae latinis,’ in Acta seminarii philologici erlangensis, 1884, III, pp. 423 ff., assigned the Visions to a later date than the Shepherd proper. A. Hilgenfeld, Hermae Pastor, 2nd ed., Leipzig, 1881 (an entirely different work from his editions of 1853, 1856, 1873), pp. 138 ff., suggested three different authors, making another division between Sim. vii and Sim. viii. Spitta, ‘Studien zum Hirten des Hermas,’ in Zur Geschiehte und Litteratur des Urchris-tentums, Göttingen, 1896, II, pp. 241–437, believes the basal work to have been Jewish, but with large Christian redactions; and in this he was followed by D. Völter, Die Visionen des Hermas, Berlin, 1900; Die apostolischen Väter, Leyden, 1904, I, pp. 170 ff.; Die älteste Predigt aus Rom, Leyden, 1908, pp. 60–69; and Völter's pupil, H. A. van Bakel, De Compositie van den Pastor Hermae, Amsterdam, 1900. H. von Soden, in the Theologische Literaturzeitung, XXII, 1897, pp. 584–587, threw out the suggestion that Hermas had passed through first a Jewish and then a Christian stage; whereas the anonymous Antiqua Mater: A Study of Christian Origins, 1887, had denied that there was anything Christian about the book. Weinel, in Hennecke's Handbuch zu den neutest. Apokryphen, commenting on Vis. v, propounded a theory not of divided authorship but of divided personality. Grosse-Brauckmann, De compositione Pastoris Hermae, Göttingen, 1910, finding no possibilities left in the way of redactors, interpolators, or separate authors, conceived the novel idea that Hermas had revised himself, the original visions having been applicable to the prophet and his immediate family but expanded for presentation to the church at large. Jean Réville, Le valeur du témoignage historique du Pasteur d'Hermas, Paris, 1900, defended the unity of authorship, but regarded the prophet's family as symbolizing the church at Rome, and the various other persons as standing for abstract qualities. K. D. Macmillan, ‘The Interpretation of the Shepherd of Hermas,’ in Biblical and Theological Studies by the Faculty of Princeton Theological Seminary, Princeton, 1912, pp. 492–543, reverted to the theory that the Shepherd is an edifying fiction. A. Link, Die Einheit des Pastor Hermae, Marburg, 1888, and still more effectively P. Baumgärtner, Die Einheit des Hermasbuchs, Tübingen, 1889, defended the unity of authorship on grounds of style and vocabulary. The fact is so self-evident that these arguments ought to have been superfluous, but the history of the interpretation of the book shows that they were not.

⁸ In the community in which the book was composed the crudity and oddity of it must have been felt for a considerable time after the author's death. Thus some fifty years later the Muratorian Canon seems to show the officials of the Roman church maintaining the same attitude as had prevailed during his life. The assertion of his late date is merely a prelude to the declaration: “And so it is fitting that he should be read; but on the other hand he cannot be read publicly in the church before the people – neither among the prophets, for their number is complete, nor among the apostles – to the end of time.” Not until it reached Alexandria and fell into the hands of the allegorists did the Shepherd really come into its own.

⁹ Vis. v, in fact, resembles a modern preface in that a part of it must have been written after the completion of the Mandates and some of the Similitudes. The addition probably begins in the middle of verse 5 with πρῶτον πάντων (repeated in Mand. i. 1).

¹⁰ C. H. Turner, ‘Marcan Usage: Notes, Critical and Exegetical, on the Second Gospel,’ Journal of Theological Studies, July 1925, p. 338. In support of the principle that “the tendency of an educated writer of ancient times would be to omit numbers,” Turner refers to the observations of the Jesuit scholar Delehaye concerning Sulpicius Severus as the biographer of St. Martin. The “incohérences de sa chronologie” are so great that by critical emendations or recombinations the date of St. Martin has been variously assigned to 395, 396, 397, 399, 400, 401, 402 (H. Delehaye, ‘Saint Martin et Sulpice Sévère,’ in Analecta Bollandiana, XXXVIII, Brussels, 1920, p. 19). Biography at that time had little interest in historical detail, and aimed rather at exalting its hero and presenting him pragmatically as a model of conduct for others to follow. Hence the “dédain de la chronologie” shown by Severus, and in general the “procédeé de ces biographes de l'antiquité à qui une date ou un calcul précis semble être une faute contre l'élégance” (p. 82). They tended to be sparing of the names of persons and places and to avoid giving exact details.

¹¹ R. Reitzenstein, Poimandres: Studien zur griechisch-ägyptischen und frühchrist-lichen Literatur, Leipzig, 1904, pp. 11–13, 33–35.

¹² Zielinski, ‘Hermes und die Hermetik,’ in Archiv für Religionswissenschaft, VIII, pp. 321–372; IX, pp. 25–60, is enthusiastic over Reitzenstein's discovery, and adds a venturesome conjecture or two of his own (VIII, p. 323, n. 1). Wilhelm Kroll, art. ‘Hermes Trismegistus,’ Pauly-Wissowa, Realencyclopädie, VIII, col. 822, line 47, expresses a belief in a common source for Hermas and the Poimandres, but gives no reasons. G. Bardy, ‘Le Pasteur d'Hermas et les Livres Hermétiques,’ Revue Biblique, N. S. VIII, 1911, pp. 391–407, is unconvinced by Reitzenstein, as are H. Weinel, in Hennecke's Handbuch zu den neutest. Apokryphen, pp. 422 f.; Dibelius, in Zeit-schrift für Kirchengeschichte, XXVI, July 1905, pp. 170–175; Krebs, Der Logos als Heiland, p. 139; Heinrici, Die Hermes-mystik und das Neue Testament, pp. 5 f.; and W. Scott, Hermetica, Oxford, 1925, II, p. 15, note 2.

¹³ No explanation offered is free from serious difficulties. Reitzenstein, Poimandres, p. 12, interprets the name as ‘Menschenhirt.’ But in the document itself there is not a word about sheep or shepherds. Also the racial idea involved in ‘shepherd of man’ or ‘shepherd of men’ ought to be expressed by a derivative of ἄνθρωπος, not of ἀνήρ. The word might be a poetical elaboration of ποιμήν, meaning something like ’ noble shepherd,’ as does the Aeschylean ποιμάνωρ made up from the same roots (Pers. 421). But unfortunately the Greek derivatives from ἀνήρ seem never to end in -ανδρης, the nearest approach being the name Ποίμανδρος, the legendary founder of Tanagra in Boeotia (Paus. ix. 20,1; Plutarch, Quaest. Gr. 37, p. 299; cf. Walter Scott, Hermetica, Oxford, 1925, II, p. 15, note 1). Hence two non-greek etymologies have been proposed: p^em^en^etre, the Coptic equivalent of the Egyptian term for ‘the witness’ (Frank Granger, ‘The Poemander of Hermes Trismegistus,’ Journal of Theological Studies, V, April 1904, pp. 395–412); and p-eime-n-re, the Coptic for ‘the knowledge of Re’ (Scott, II, p. 16, reporting the suggestion as coming from F. Ll. Griffith). The insertion of the δ would be euphonic merely, and the recurring appositive ὁ τῆς αὐθεντίας νοῦς would be a fairly apt translation of the Coptic. This is admirable until one considers the articles. In the Greek document it is always ‘the Poimandres’ (except the one vocative in § 16, ‘O Poimandres’). But in Coptic the first syllable ‘p’ is itself the article. Thus a curious reduplication results, as if one should say ‘the hoi polloi’ (as the imperfectly educated sometimes do) or ‘the la belle France.’ To be sure, Fulgentius the mythographer did just this with the Greek when he spoke of “the book of Opimandres” (Helm's edition, p. 26, line 18). But the author of the Hermetic document appears to have been better educated than Fulgentius, and also by hypothesis he knew the derivation of the word he was using.

The author of Corp. Herm. xiii. 19 apparently connected the name with the idea of shepherding, ποιμαίνει ὁ νοῦς; but since this thirteenth document is unquestionably later than the first (cf. Scott, II, p. 373), it gives no necessary indication as to what the earlier writer had in mind. Hermas – if one accept his testimony at all – seems to favor a Greek derivation. He sets the word ἀνήρ in the very forefront of his description in Vis. v. 1, as if he understood the name ‘Poimandres’ to mean ‘shepherd-man,’ like ποιμνωρ. But what Hermas understood throws very little light on what the original author of Corp. Herm. i may have intended.

¹⁴ One may even suspect that the use of the Poimandres by Hermas (if this be admitted as likely) had something to do with the case; cf. the supposition of Burkitt (note 27, below) as to how the Manichees came to use the Shepherd. The genitive of Hermas is identical with that of Hermes, and ὁ Ποιμήν τοῦ Ἑρμαῦ and ὁ Ποιμάνδρης τοῦ Ἑρμοῦ having become confused, the latter might easily be taken to refer to the god. But this is all highly conjectural.

¹⁵ Cf. Vis. iii. 3,1; 3, 2; 6,5; 7, 5; iv. 3,1; Sim. v. 4, 2; 5,1. “Accurately” is his great word (ἀκριβῶς and its derivatives); cf. Vis. iii. 10, 10; Mand. iii. 4; iv. 2, 3; 3, 3; 3, 7; Sim. ix. 1, 3; 5, 5; 13, 6.

¹⁶ The historical situation under Trajan is vividly outlined by C. Bigg, Origins of Christianity, Oxford, 1909, pp. 84–98.

¹⁷ Cf. Sim. viii. 3, 3 for the mention of Michael, and Sim. ix. 11, 15, and passim, for the personified Virtues. W. Lueken, Michael: Eine Darstellung und Vergleichung der jüdischen und der morgenländisch-christlichen Tradition vom Erzengel Michael, Göttingen, 1898, gives a valuable exposition of the religious importance of Michael in post-prophetic Judaism. The early Yahweh had been a nationalistic champion, but the prophets transformed him into a universal deity. Thereupon the old particularism transferred itself to the principal angel, who from the time of the Book of Daniel onward appears as the heavenly guardian of the Jewish nation. When therefore the Christians arose to claim that they were the true Israel, it was not difficult for them to identify their Christ with Michael.

¹⁸ Nathan A. Harvey, Imaginary Playmates and Other Mental Phenomena of Children, Ypsilanti, Michigan, 1918, p. 16, after investigating 109 cases, found only three instances in which the phenomenon persisted into adult life. The matter is treated also in W. S. Taylor, Readings in Abnormal Psychology, New York, 1926, p. 601.

¹⁹ Several papyrus fragments of Hermas are known, but the collection belonging to the University of Michigan has recently yielded a papyrus codex extending from Sim. ii. 9 to Sim. ix. 5, 1. This is to be edited by Professor Campbell Bonner, who gives an account of the manuscript, with specimens of the peculiar readings, in ‘A Papyrus Codex of the Shepherd of Hermas,’ Harvard Theological Review, vol. XVIII, 1925, pp. 115–127. See also the note by Lake in the same volume, pp. 279 f.

²⁰ The testimonia are most conveniently collected in Harnack, Geschichte der altchristlichen Litteratur, Part I (Leipzig, 1893), pp. 49–58.

²¹ One of the Latin manuscripts, the Codex Palatinus, leaves an extra space and inserts a large ‘Amen’ at the end of Similitude viii, and puts another ‘Amen’ after Similitude ix. Harnack, in his commentary, p. 69, expresses the opinion that this goes back to the original copy, and that it marks certain of the stages in the publication of the work. P. Baumgärtner, Die Einheit des Hermasbuchs, Tübingen, 1889, p. 38, thinks that the ‘Amens’ indicate new authors for Similitude ix and Similitude x. This is resting a good deal upon a mere pious ejaculation. There are undoubtedly new stages in the composition at these points, but not in the publication, much less in the authorship. Perhaps some scribe wrote the ‘Amens’ in glad relief after he had worked his way through the arid wastes of Similitude viii and Similitude ix. There is no trace of them in the Michigan papyrus.

²² Something may be inferred also from Ignatius, Rom. 5, 2, in which he speaks as if martyrdoms like his own had been a not uncommon occurrence at Rome: “May I have joy of the beasts that have been prepared for me; and I pray that I may find them prompt. I will even entice them to devour me promptly – not as they have done to some, whom they would not touch through fear.” This letter may be as early as the letter of Pliny, though it is usually assigned to a date later in the reign of Trajan.

²³ Cited and discussed by Harnack, Geschichte der altchristlichen Litteratur, II. 1 (Leipzig, 1897), p. 266, note 2. Hilgenfeld firmly believed in the dependence of Hermas on the teachings of the Elkesaites, and in both his earlier and his later editions (Die apostolischen Väter, Leipzig, 1866, pp. 151–167, and Hermae Pastor, Leipzig, 1881, pp. 230 ff.) he printed the notices of this heresy as found in Hippolytus, Origen, and Epi-phanius.

²⁴ Such a result accords with the opinions of the most careful critics, in so far as they express themselves at all, respecting the extent of the prophet's career. For example, C. Bigg, Origins of Christianity, Oxford, 1909, p. 73, says: “We may suppose that three or four years had elapsed.” Even Harnack, who has spread the chronology of Hermas out over a longer period in order to date the final publication ca. 140, is manifestly disturbed at the outcome of his own reasoning. In the Chronologie (Geschichte der altchristlichen Litteratur, II. 1, p. 266) he observes: “The previous stages, including Visions i-iv, extend over a considerable time – not more, however, than twenty or twenty-five years (it might very well be less).”

²⁵ So Tertullian, De pudicitia, 10: “ab omni concilio ecclesiarum etiam vestrarum [i. e., Catholic as opposed to Montanist] inter apocrypha et falsa iudicaretur.”

²⁶ If, as Tertullian indicates, it had been widely condemned, that means that it had also been widely read in Africa. For its use in Gaul, see Irenaeus, Contra haeres. iv. 20, citing Mand. i. 1. Eusebius, H. E., v. 8, 7, assumes that Irenaeus is here ascribing canonicity to Hermas, but this is not quite certain. “Scriptura” may mean merely ‘writing.’

²⁷ F. W. K. Müller, “Eine Hermas-stelle in manichäischer Version,’ Sitzungsberichte, Berlin Academy, 1905, pp. 1077–1083, published in parallel columns (1) the Persian text of the fragment, M. 97, (2) the translation, and (3) Weinel's German translation of the pertinent portions of Sim. ix. 4, 7; 6, 1; 9, 4 f.; 18, 5; 19–25. The identity of the passage is beyond question, but it seems more like a free, condensed summary than an exact translation. At some time between the third century and the tenth the religion of Mani spread into Chinese Turkestan, and from the neighborhood of Turfan have been recovered several hundred Manichaean manuscripts, all in fragmentary condition. Why one of these should contain the excerpts from the Shepherd is obscure. F. C. Burkitt, Religion of the Manichees, Cambridge, 1925, pp. 95–97, suggests that some of the western missionaries of the sect may have picked up a copy of the revelations of Hermas and confounded them with the writings ascribed to Hermes Trismegistus, who is known to have been held in honor by certain of the Manichees.