The Pastoral Epistles: A Commentary on the Greek Text

By STANLEY E PORTER

Leading New Testament scholar Stanley Porter offers a comprehensive commentary on the Pastoral Epistles that features rigorous biblical scholarship and emphasizes Greek language and linguistics.

This book breaks new ground in its interpretation of these controversial letters by focusing on the Greek text and utilizing a linguistically informed exegetical method that draws on various elements in contemporary language study. Porter pays attention to the overall argument of the Pastoral Epistles while also analyzing word meanings and grammatical structures to tease out the textual meaning. Porter addresses major exegetical issues that arise in numerous highly disputed passages and--while attentive to the history of scholarship on First Timothy, Second Timothy, and Titus--often takes untraditional or innovative positions to blaze a new path forward rather than adopt settled answers.

This commentary will appeal to professors, students, and scholars of the New Testament.

• Language: English
• Publisher: Baker Publishing Group
• Release date: Oct 17, 2023
• ISBN: 9781493436880

Book preview

Porter writes a different kind of commentary for the Pastoral Epistles, one devoted to an accessible linguistic interpretation of the Greek that helps the reader understand the language of the texts and their situational context. While focusing on insights at the clause level, he also sheds new light by bringing linguistic theory and the data of the texts to bear on the critical issues and the notorious interpretive problems of these controversial and intriguing letters.

—Cynthia Long Westfall, McMaster Divinity College

This is the most thorough commentary in English since the 1999 commentary by Marshall. The introduction is itself a reason to buy this commentary. Porter’s vast knowledge of linguistics and primary and secondary sources is on display in this magnificent commentary, which I suspect will become the standard for the exegesis of the Greek text.

—Osvaldo Padilla, Beeson Divinity School

This is a much-needed book on a contentious group of writings. Porter sheds fresh light on the Pastoral Epistles and the issues surrounding them, especially authorship, by applying a consistent and rigorous linguistic model (Systemic Functional Linguistics) and placing emphasis where it belongs, on the Greek text. No one will want to study these books without recourse to Porter’s meticulous, judicious, and thought-provoking work. Get this book, and refer to it often.

—David L. Mathewson, Denver Seminary

This is not just one commentary among many but a unique commentary on the Greek text of the Pastoral Letters. It is intensive and detailed (clause by clause, not word for word, explaining the meanings in context), consistent (based on the linguistic method, Formal Systemic Functional Grammar, explained in the prolegomenon), thorough and encyclopedic (discussing the opinions of other commentators), and extensive (about 1,000 pages). Its language is accessible to the average reader, avoiding technicalities as much as possible. This is a commentary full of original interpretations that presents a new and comprehensive reading of the Pastoral Letters.

—Jesús Peláez, University of Córdoba, Spain

If Mount Rushmore honored great and recent English-language commentaries on the Pastorals, alongside the works of I. Howard Marshall, Philip Towner, and Gerald Bray would now appear this offering by Stanley E. Porter. Through rigorous application of Formal Systemic Functional Grammar, Porter sets forth a reading that will provoke and inform constructive discussion for years to come, not least because of conclusive arguments advanced for Pauline authorship.

—Robert W. Yarbrough, Covenant Theological Seminary

© 2023 by Stanley E. Porter

Published by Baker Academic

a division of Baker Publishing Group

PO Box 6287, Grand Rapids, MI 49516-6287

www.bakeracademic.com

Ebook edition created 2023

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means—for example, electronic, photocopy, recording—without the prior written permission of the publisher. The only exception is brief quotations in printed reviews.

Library of Congress Cataloging-in-Publication Data is on file at the Library of Congress, Washington, DC.

ISBN 978-1-4934-3688-0

Unless otherwise indicated, all Scripture quotations are the author’s translation.

Baker Publishing Group publications use paper produced from sustainable forestry practices and post-consumer waste whenever possible.

Contents

Cover

Endorsements    i

Title Page    iii

Copyright Page    iv

Preface    vii

List of Abbreviations    ix

Prolegomenon    1

Introduction to the Pastoral Epistles    19

1 Timothy

I. Letter Opening (1:1–2)    109

II. Thanksgiving    125

III. Body: Behaving as a Leader of God’s Church (1:3–4:16)    127

A. Formal Opening: Refuting the False Teachers (1:3–11)    129

B. Paul’s Faithfulness to His Calling (1:12–20)    164

C. Instructions for Orderly Conduct inside and outside the Church (2:1–3:16)    201

D. Timothy’s Role in the Church (4:1–16)    328

IV. Paraenesis (5:1–6:19)    374

A. Household Duties (5:1–6:2)    376

B. The Return to False Doctrine versus Godliness (6:3–10)    444

C. Pursue What Is Righteous (6:11–19)    466

V. Letter Closing (6:20–21)    493

2 Timothy

I. Letter Opening (1:1–2)    503

II. Thanksgiving (1:3–5)    509

vi III. Body: Serving Christ (1:6–4:18)    519

A. Formal Opening: God Gives a Spirit of Power (1:6–14)    521

B. Faithful and Unfaithful Friends (1:15–18)    545

C. Endurance Is for Long-Term Service (2:1–13)    556

D. False Teaching Is to Be Resisted (2:14–26)    582

E. Eschatological Climax: Evil Behavior Brings Divine Rejection (3:1–9)    616

F. Apostolic Charge (3:10–4:18)    631

IV. Paraenesis    693

V. Letter Closing (4:19–22)    695

Titus

I. Letter Opening (1:1–4)    707

II. Thanksgiving    725

III. Body: Leading the Church (1:5–16)    727

IV. Paraenesis (2:1–3:14)    763

A. Teaching Sound Doctrine (2:1–15)    765

B. Living under Rulers, Authorities, and Others (3:1–7)    808

C. Avoiding Divisiveness (3:8–11)    832

D. Personal Comments (3:12–14)    842

V. Letter Closing (3:15)    852

Works Cited    859

Index of Modern Authors    927

Index of Scripture and Other Ancient Sources    949

Back Cover    970

Preface

A work such as this is a communal effort. I want to thank Moisés Silva for first inviting me so many years ago to write a commentary on the Pastoral Epistles. The finishing of it has been a long time in coming, and it has no doubt gone in directions that Moisés would never have anticipated. The commentary itself has become something that admittedly does not comfortably fit within the rather predictable commentary style that is typically found these days. I also wish to thank Bryan Dyer and Baker Academic for the opportunity to publish this commentary and their full support of it as an attempt—even if an incomplete and iconoclastic one—to write a different kind of commentary. On the one hand, its linguistic orientation will be consternating to many readers; on the other hand, its attention to matters of language will remind some of commentaries from previous and sometimes long-distant eras.

I thank my many scholarly collaborators through the years whom I have had the pleasure of working with on various projects. Many, but not all, of them are listed in various forms among the Works Cited. One of the great pleasures of my scholarly career has been the opportunity to work with many exceptional scholars, often in areas of new and exciting exploration. Much of that work is evident in this volume.

I also thank several people who read through this manuscript and made comments on it at various times and in various forms. In particular, I thank Robert Yarbrough for his thorough reading, which helped to improve the manuscript at several places, even if he was not in agreement with many of the things that I argue. After I had written my section on 1 Tim. 2:8–15, I had the opportunity to read the abstract of a paper by David Warren that helped me to refine my arguments further. Drake Williams also provided some useful last-minute bibliography.

I also thank numerous students who have engaged in discussion of the Pastoral Epistles over the years, most recently Moises Zumaeta and Yan Ma, who have also provided some useful comments on the scholarship of others. I thank several graduate assistants who have immeasurably helped me to complete this task. Through my thirty and more years of teaching, I have had the tremendous opportunity to work with exceptional students, many of whom have gone on to their own significant careers, sometimes in biblical studies but other times in fields equally important. I am sure that I have forgotten some of those who have worked diligently on this project, but several of them must be mentioned. These include David I. Yoon, who assembled much of the initial bibliography for this volume and read several of the sections with his usual critical and constructive eye; John J. H. Lee, who performed some last-minute bibliographical searching; and Jihyung Kim, who has done just about everything imaginable and more. Thank you, Ji, for reading through the entire manuscript, for your diligence in searching for and re-searching for sources, and for your general support and efficiency throughout the project. I am deeply thankful.

I am very thankful for all the work that others have done to help ensure that this commentary paves new ground. I especially want to thank my several editors at Baker Academic, who have thoroughly and constructively worked through this manuscript and saved me from many errors. In particular, I wish to thank Wells Turner for his incredible attention to detail, engendered in his entire team. There are bound to be remaining errors and mistakes, for which I apologize. This commentary has become a much more complex endeavor than most can even imagine, and mistakes are inevitable. However, I think that it also does some things that are unusual and possibly even unique in commentary writing; the risk and benefit of breaking out of the traditional mold, so far as that is possible, are worth enduring some difficulties along the way.

Last, I thank my wonderful wife, Wendy, for making life the great adventure that it is, and especially for her tremendous love and support during this dreaded COVID-19 pandemic. Wendy, thank you for keeping everything going and doing everything you could to keep us safe and well when others about us had gone absolutely mad.

I wish to dedicate this volume to one of my first teachers in New Testament, Walter Liefeld. This is probably not the commentary that he envisioned, but he was an inspiration in its conception and execution.

Abbreviations

General and Bibliographic

Prolegomenon

Here I will offer the reader an orientation to some of the features of this commentary that should enable the reader to use it in a more productive way. Those commentaries that provide such an orientational feature often say very little (e.g., Fiore 2007: 5–7, who offers virtually no guidance such as is found below; L. Johnson 2001: 98–99, who offers three brief comments; and R. Yarbrough 2018: 67–69; cf. Himes 2010: 92, who asks for comments on verbal aspect; see below). Some readers may find it confusing that I have written what appears, at least at first glance, to be a rather traditional commentary, since I have criticized conventional commentary writing on various occasions (e.g., S. Porter 2013a; 2013e; 2016c). My major criticism against conventional commentary writing is that most of them offer little that is new in interpretation, to a large extent because they have departed from the major purpose of a commentary on an ancient text: help in understanding the language of the text and thereby its situational context. To do this, commentaries must be primarily attentive to linguistic understanding rather than focused on any number of other purposes that distract from this primary issue. I also criticize commentaries for attempting to be compendia of knowledge, especially theological and related knowledge, rather than providing focused treatment of the text of the individual biblical book. In this commentary, I attempt to address these problems by providing a commentary on the Greek text of the Pastoral Epistles, described through the framework of Systemic Functional Linguistics as I understand it (a framework on which I have written at length; see below and Works Cited). Whatever else this commentary may provide, I have tried to produce a consistent, thorough, and informative commentary on the Greek text of the Pastoral Epistles by means of a form of Systemic Functional Linguistics—specifically, Formal Systemic Functional Grammar—that I believe helps to enlighten our understanding of ancient Greek.

I will say more about this linguistic framework below, but before I do so, I should explain the process that resulted in this commentary. I was originally asked to write this commentary because, according to the person who asked me, I had the right answer on the question of authorship, and he apparently wanted someone to defend that view of authorship. The right answer was to defend Pauline authorship of the Pastoral Epistles. I held to that position at the time that I accepted the offer to write the commentary and I still hold to that position. I have written on this position several times, most recently in a volume on Paul where I introduce all thirteen of his letters, all of which I consider authentic (S. Porter 1995; 1996a; 2016a). I will say more about the topic of authorship in the introduction to the Pastoral Epistles that follows this prolegomenon.

When I began writing this commentary—or rather, these commentaries—on three different NT letters, I surrounded myself with roughly thirty other commentaries so that I could avail myself of the breadth and diversity of scholarship on these letters. I purposed to use these commentaries to inform my own commentary writing. I was sadly disappointed with the result. I rarely found genuinely new and insightful comments on the Greek text and found a variety of other less helpful commentary practices, such as extended comments on various translations, comments mostly on a selected English text, or comments that drew out a specific theological issue with which the commentary was concerned. What I found more than anything else was a large amount of repetition among the commentaries. I must confess that I soon began whittling down the number of commentaries that I regularly consulted until I was left with two commentaries as my constant companions, and I have benefited immensely from both of them. I am not denying that there are some valuable comments in other commentaries; there no doubt are, as I have tried to make clear in the commentary itself. But the effort required to read quantities of extraneous (and repeated) material was counterproductive to keeping my own perspective and even sanity as I worked through the text.

The two commentaries that I ended up constantly referring to were those by George Knight (1992) and I. Howard Marshall and Philip H. Towner (1999). Through their use, I found that I was generally exposed to many if not most of the major issues regarding understanding the Greek text (mostly through Knight, but also through Marshall/Towner) and to most of the major critical scholarly debates over exegetical issues within the books (mostly through Marshall/Towner, but also through Knight). As a result, I freely admit—and want to make it plain here, lest there be any misunderstanding—that I have looked to Knight and especially Marshall/Towner as repositories of references to the variety of scholarly opinions, including the readings of various translations and ancient sources, and have drawn upon those lists of opinions and references in my commentary. Thus, those who would compare these two commentaries with mine will find that I have often referred to similar secondary literature and even primary sources that they cite. I have, so far as possible (especially during the COVID-19 pandemic), verified these references, so that the result is, I believe, a faithful representation of the state of scholarship. I, like other commentary writers, have also plundered the Egyptians by using other commentaries as sources of references as well, again trying wherever possible to verify the references.

I have also added numerous references of my own that make the scholarly apparatus even more thorough and comprehensive in many ways than that of these other commentaries, especially where I have cited later commentaries and other authors. What I have just described is, in fact, what I have found that probably some other scholars do in commentary writing: using the work of their predecessors as the basis for their own, at least so far as access to secondary literature is concerned. Let me also make clear, however, that I have looked to Knight and Marshall/Towner (as well as other commentators and scholarly works) to provide a check on my own work on the Greek and to ensure that, as much as is reasonably possible, I cover the range of scholarly debate on any given topic.

I examine the Greek text of the Pastoral Epistles from a linguistic perspective, but this is not a full-blown linguistic commentary. I attempt to minimize the jargon and put the description and explanation into language that NT scholars without linguistics will readily be able to understand, although I admit that it is difficult not to use some linguistic terminology. However, I think it is important to make clear the linguistic framework that I use. I find it highly problematic when people use the term linguistics without defining what they mean or use an eclectic model that haphazardly seems to draw upon elements from various frameworks without making clear why they do so. This is not the place to outline a general theory of linguistics or to exposit in detail the entire framework I utilize. Nevertheless, I think it is important to mention some of the major elements of the framework that rest behind my specific linguistic descriptions within the commentary.

The linguistic framework that I utilize is Systemic Functional Linguistics.1 Systemic Functional Linguistics, and more particularly Formal Systemic Functional Grammar, is a system-based functional linguistic model that connects socially grounded meanings with instances of language usage. Systemic Functional Linguistics relies on defining and examining various theoretical strata that connect context to expression. Each of these strata—context of culture and context of situation (which are nonlinguistic), semantics and lexicogrammar (content), and phonology and graphology (expression)—is system driven. In this commentary, I concentrate on the content level, dealing with semantics and lexicogrammar, although the astute reader will notice that I also make use of a level that I have called discourse semantics to handle the semantics/pragmatics interface (see S. Porter 2016f). Systemic Functional Linguistics models meaning potential as system networks, in which meaning choices are realized as systems. Also, Systemic Functional Linguistics utilizes a rank scale to differentiate levels of structure (syntagmatic relations) of language. Cutting across these strata are three major metafunctions: the ideational (concerned with the what of language), the interpersonal (concerned with the who of language), and the textual (concerned with the how of language). This has already been demonstrated in a previous commentary (S. Porter 2015a).

In Formal Systemic Functional Grammar, there is a strong correlation between form and function: the forms of the language express their semantic functions. I do not emphasize the semantic/pragmatic divide because I hold to a view of grammatical and lexical monosemy, even though monosemy is not a distinctively systemic functional feature. Monosemy is the principle, or perhaps the orientation or predisposition, of seeing singular rather than multiple meanings for any linguistic element. Lexical monosemy is the view that individual lexemes have an abstract meaning based on the range of their uses as modulated by context.2 Charles Ruhl (1989) refers to what he calls the monosemic bias, in which one may choose to view the lexicon as either polysemous or monosemous.

The monosemic bias, I believe, provides a more productive way of considering the lexicon of a language since users associate words with broad and abstract concepts that may or may not be reflected in individual instances of translation. This semantic meaning is then constrained by various factors in the context, both lexical and grammatical, so that one arrives at a contextualized meaning. In other words, the meaning moves from the broad and general in the abstract to the particular in a given instance. This overcomes the typical problem of the semantics/pragmatics divide where the semantic meaning is usually seen as a specific meaning or a typical meaning that is often difficult to see expressed pragmatically. The monosemic bias also overcomes the idea that semantics is the invariable meaning left after pragmatics is removed. This notion makes pragmatics into something larger or outside or different from semantics. If we reverse the notion, we see that context (what some may see as pragmatics) modulates and narrows the potential of the semantics, and thus one does not need to distinguish between the two. The semantic meaning is always present (otherwise we are speaking of homophones or homographs), even if constrained.

In this commentary, as a result, if I engage in any word studies, as is usually seen in commentaries, I have made a mistake. I make lexical comments, but they are of the order of establishing the abstract meaning that I then see constrained by context (including both lexical and grammatical elements). I do not often refer to such theological lexica as TDNT, where heaps of theological meanings found in various contexts are loaded onto individual lexemes (see J. Barr 1961). I have also found that BDAG, though always of use, is not always a perfect tool because its principles of classification are greatly mixed and reflect mostly pragmatic instances rather than semantic meanings (see S. Porter 2015b: 61–80; see also S. Porter 2019b, for a review of recent works on lexical semantics; cf. J. A. Lee 2003; Peláez/Mateos 2018). I instead have primarily relied on the Louw-Nida semantic domain lexicon (1988) and Montanari (2015), although both reflect polysemy rather than monosemy (see S. Porter 1996b: 49–74). However, I have found that both are useful, especially Montanari, as a means of determining the abstract semantic sense of a given lexeme, which may then be modulated or metaphorically extended or otherwise shaped by context. I try to capture these abstract and specific meanings in my discussion of individual words, although I make every attempt to avoid making this a typical (sadly) word-for-word commentary.

I also use a similar grammatical monosemy, first seen in my monograph on verbal aspect and then in most of my subsequent linguistic work (e.g., S. Porter 1989; 2016f; S. Porter/Pitts 2009; see Frey 1998: 51, who clearly labels my approach monosemic). Grammatical monosemy is similar to lexical monosemy in positing that grammatical features also have abstract semantics that are modulated or constrained by contextual features, including grammatical environments. This also overcomes the semantics/pragmatics divide by not relying on overly specific semantic meanings that cannot be aligned with pragmatic usage and not relying on typological meanings that are representative of some but far from all uses. As I mentioned above, semantics is not concerned with the irreducible core after pragmatics is removed, but with the abstract concept that is shaped into specific contextual meanings. The accumulative model is often used, in which pragmatics is seen as an agglutination of various features that then must be canceled, with cancelability of pragmatic features leaving the semantic core.3 The movement is just the opposite. Rather than moving from one specific (semantic) meaning to another even larger one (pragmatic), the movement is from the abstract (meaning of the form) to the modulated (meaning in context), or large to small meaning.

Numerous examples may be cited. These include the verbal aspects, where the aspect associated with the aorist tense form, what I am content to call perfective aspect, is semantically present in every use of the aorist, even if it is modulated by its individual context for such a feature as temporal location (see S. Porter 1989). The same may be said of the verbal attitudes, where the attitude associated with a given mood form, such as the subjunctive with its semantics of projection, is found in every use of the subjunctive, even if it is modulated by its individual context to a projective statement (hence accounting for a negated aorist subjunctive as prohibition; see S. Porter 2016f).

Similarly, a semantic meaning of every Greek case is found in all its pragmatic uses (see S. Porter 1994: 80–100, following Louw 1966; and S. Porter 2021), even as there is an important distinction between the nominative and the oblique cases, and then among the oblique cases (with oblique cases used in nominative groups as types of modifiers). The vocative case is a limited or partial case, found only in restricted instances in the singular (not the plural). Some commentators confuse the nominative of address in the plural with the vocative (see S. Porter 1994: 87–88; S. Porter/M. O’Donnell 2006).

I approached the writing of this commentary from a standpoint similar to that in my commentary on Romans, considering it a form of discourse register analysis (S. Porter 2015a: 24). Register studies have become very important for the study of the Pastoral Epistles (see the introduction to the Pastoral Epistles) because register variation is concerned with variation according to use.4 The notion of register provides a means of accounting for linguistic variation on the basis of use and context, rather than on the basis of variation in author, and has been used to that effect in study of the Pastoral Epistles (see Libby 2016; M. O’Donnell 1999; 2005: 85–101; Pitts 2013b; Pitts/Tyra 2017; S. Porter 2000b; van Nes 2018c). In this commentary, register is a term to be taken as roughly equivalent to genre. I believe that the ancient letter provides a relatively closed register, whose semantic structure is driven by the epistolary form, but within which are open variations, what one might call subregisters, such as the petition or testamentary letter or the like. I use the letter form as a primary means of establishing functional places for interpretation within the letter as a register; then I develop the argument within those epistolary locations. There have already been several other register analyses of Pauline letters worth noting (e.g., Land 2015; Reed 1997; Yoon 2019), besides examples of discourse analysis (e.g., Allen 2020 using Relevance Theory).

The major focus of my Greek grammatical discussion in this commentary is on the clause, since the clause is the major semantic and structural unit in the Greek language—what Michael Halliday calls clause as message (2014: 88). I focus on the material associated with the clause, even if I often go beyond the clause or beneath the clause. Much of the framework below is utilized in the OpenText.org syntactical display of the Greek NT, which has been made for both clause and word-group ranks (see Land/Pang 2017 on OpenText.org). Here I offer the most basic sketch of my description of language to make clear some of the terminology I use throughout the commentary.

At the level of the clause, I utilize a single functional analysis to capture the ideational and interpersonal semantics by identifying four clausal functional components: the subject (S), predicator (P), complement (C), and adjunct (A). These labels identify functions within the clause. The subject may or may not be grammatically expressed as a clausal component in Greek. The subject varies in its function as active agent or medium of the clause according to the voice/causality of the verb (see below), but its function is in the causality system. The predicator indicates the process of the clause and semantically expresses the aspect/tense-system, attitude/mood, and causality/voice (with the first of each pair indicating the functional system and the second the formal expression) as well as person and number. The major systems are discussed below. The complement, which can be direct or indirect, is the recipient (however that may be defined) of the action of the process. These three components—subject, predicator, and complement—are the major clausal components of the Greek clause. The adjunct is an optional component that indicates the circumstances of the clause and may refer to such circumstances as time, place, frequency, manner, or the like.

There are probabilistic or usual ways in which these elements may be expressed (e.g., subject by a nominative element, predicator by a verb, complement by a range of elements, and adjunct by another group of elements, often including prepositional phrases or [embedded] participle clauses). This clausal analysis may be used for primary or secondary clauses, and even for participle and infinitive clauses with some constrictions (e.g., participle clauses do not have grammatical subjects, only relational agents/mediums).

The NT Greek clausal component ordering has been the subject of limited discussion (see Pitts 2013a, who basically confirms the word and constituent order patterns presented in S. Porter 1993d; 1994: 286–97; 2015b: 347–62; cf. Tan 2009, who provides similar information). There are some who still use the terminology of VSO (verb, subject, object) or SVO to discuss Greek, but this assumes a grammatically explicit subject (whether full form, such as a noun, or reduced form, such as a pronoun), which many Greek clauses simply do not formally express (they have an implied subject with verbal morphology). The three most common clausal types, at least in primary clauses, appear to be P (predicator), PC (predicator + complement), and then CP (complement + predicator) and SP (subject + predicator). The patterns that vary from this are creating emphasis on the individual component that is displaced, usually to the front of the clause to thematize it.

The subject, complement, and adjunct are filled by word groups of various types, such as nominal word groups formed around a noun or similar entity as head term, often used within the subject or complement, and adverb or prepositional groups within the adjunct. The verbal group fills the predicator, typically as a single word, a finite verb, but as a finite and predicator with participle in a periphrastic construction. Rank shifting occurs when a larger ranked element fills a lower one. This is seen, for example, when a participle clause (a clause with a participle as its predicator) may fill the adjunct component and function as what is usually called an adverbial participle (such constructions are referred to as [embedded] adjunctive participle clauses, even if their adverbial function is described), or when a participle clause serves another modifying function (see below). Infinitive clauses may fill a variety of slots also, with the infinitive as the predicator of the infinitive clause even if it fills another clausal component. Word groups and clauses function at different ranks on the rank scale and are subject to different ordering patterns within word groups (see below on word groups) and clauses (see above), as well as in clause complexes (see below).

At this point, let me make clear my stance on some of the major issues regarding the Greek verb. The Greek verb is often spoken of in relationship to tense, aspect, and mood (TAM, or sometimes TMA), although causality/voice should also be a component (thus TAMV). I will say something about each of these. The Greek verbal system is aspectual, with the aspects realized by the so-called tense forms. The aspect system functions within the ideational metafunction. This is what I mean by grammatical aspect, that the aspects are realized by synthetic grammatical categories. This stands in contrast to other methods of classification and analysis, such as Aktionsart, procedural aspect, lexical aspect, process types, or any of the other categories sometimes used, which are at times based on philosophical categories, typological classifications, or even temporal categories, with their level of articulation varying from the word to the clause and beyond.5

As I have argued elsewhere, I believe that the Greek verbal system is aspect prominent and nontemporal, by which I mean that the so-called tense forms in Greek grammatically express verbal aspect—the author’s subjective conception of the process—and not temporal values. This does not mean that Greek language users could not indicate temporal relations; they did so by way of other contextual discourse features, including deictic indicators, event sequencing (see Reber 2016), text-type, and so on, not the tense forms themselves (perhaps the majority of the world’s languages are tenseless; see Lin 2012: 669). This view is roughly held by not only S. Porter (1989; 1994; 2015b: 159–94) but also McKay (1994), Decker (2001), C. Campbell (2007; 2008), Mathewson (2010), and Huffman (2014), among others. The more time-bound view of aspect is held by Fanning (1990) and Olsen (1994). Many others have expressed their opinion on this topic (e.g., Brookins 2018, among others).

Trying to define the different aspects has proved to be difficult since some authors define them in typological terms, others in modal terms, and some in temporal terms—all of which miss the point of the meaning of grammatical verbal aspect—but most are content with the language of perfective versus imperfective for the aorist versus the present/imperfective opposition. I will use such labels here, recognizing that the label does not necessarily capture the full semantic value of the individual aspects. I believe, however, that there are three verbal aspects: (1) the perfective aspect, realized by the aorist tense form for a process seen to be complete; (2) the imperfective aspect, realized by the present tense form and the imperfect for a process seen to be progressive; and (3) the stative aspect, realized by the perfect tense form and the pluperfect for a process seen to represent a state of affairs.6 The imperfect and pluperfect are subcategories within their larger respective aspectual categories. The perfective aspect is the relatively unmarked (assumed) aspect: it is the default choice and hence is often used to carry the narrative line in discourse. The imperfective aspect is marked (overtly signaled) in relation to the perfective and hence is used to highlight an action or process in narrative or in exposition against the background of the perfective. The stative aspect has been the subject of continuing discussion (see C. Campbell, Fanning, and S. Porter 2021; S. Porter 2015b: 195–215) but is the most marked of the aspects and indicates a state of being or affairs of the agent of the process.

Despite fairly widespread agreement among those who have studied the issue that Greek is aspect prominent, there are still differences of opinion on some matters. For example, I believe that there are three verbal aspects, not two, as some have posited: the perfective, the imperfective, and the stative.7 These are organized by means of two sets of binary systemic oppositions. There are some who hold to a perfective versus imperfective opposition with the perfect tense form as perfective (Fanning 1990: 290–309) or with the perfect tense form as heightened proximity (C. Campbell 2007: 161–212; 2011). There are other views as well (e.g., Andrason/Locatell 2016; Crellin 2016a; 2016b). There is also some dispute over the temporality and the aspectuality of the future form. The future form, however, is nonaspectual and nontemporal (vague), grammaticalizing the semantic feature of expectation; hence its use in context is compatible with both indicative and nonindicative forms (S. Porter 1989: 403–39; see Decker 2001; Mathewson 2010; contra Fanning 1990 and Olsen 1994, who take a temporal view; C. Campbell 2007, who takes it as perfective and future-referring; and Huffman 2014, who takes it as perfective and proximate).

The attitude system is realized by the Greek mood system, which falls within the interpersonal metafunction. There has been much less research on the Greek mood system (see Mathewson 2021; S. Porter 1989: 163–81; 1994: 50–61; 2016f; 2018; S. Porter/M. O’Donnell 2001; cf. Mathewson/Emig 2016: 160–91). The Greek attitude system is concerned with the language user’s perspective on the action’s relationship to reality. The attitudinal system is reflective of clause types, with their respective mood forms.

We find two major systems within the attitude system: assertive versus nonassertive. The assertive attitude, realized by the indicative mood, is used to make declarations or ask questions, not to reflect reality. The nonassertive attitude is either directive, used to express commands (realized by the imperative mood); or projective, used to project processes without contingency (subjunctive mood) or with contingency (optative mood). This results in five major exchange roles of the attitude system: giving, projecting, wishing, demanding, and inquiring (see S. Porter 2016f: 27, 29). In the commentary, I abbreviate my comments regarding mood; nevertheless, the system is in the background. Since the Greek verbal system is nontemporal, none of the aspects or the attitudes indicate temporal location, but they have temporal values only in the context of the larger discourse. Their semantics is more concerned with making statements about perceived realities, such that one might wish to make assertions or direct others to do or project for the sake of discussion. There is a hierarchy of semantic weight for the nonassertive attitudes.

The directive, used for commands and realized by the imperative, is the least marked and is concerned with attempting to direct others. Differentiation between the aspects in commands and prohibitions is based on aspect, not whether actions are complete or continuing or on the basis of frequency (on commands and prohibitions in Greek, see Dvorak 2018; Huffman 2014; Mathewson 1996; 2021; S. Porter 1989: 335–60; 1994: 220–29; 2016f; 2018; for other perspectives, see M. Aubrey 2016; Fantin 2010; 2018; Himes 2010, who does not refer to S. Porter 1989). The projective is more marked and projects a conception of reality; the projective with contingency is the most marked and projects with a contingent view to accomplishment. Because of a variety of factors, not least its constrained paradigm of forms that indicate its marginal status within the Greek verbal system as being nonaspectual, the future form (I prefer not to call it a tense form) indicates the semantic feature of expectation. It is aspectually vague: it does not realize an aspectual value but conveys the notion of expecting a process to come about. Therefore, it is used in both assertive and nonassertive contexts. Context may indicate that the clause has future reference, but that reference is based on the clausal context, not the future form itself. Participles and infinitives realize many semantic features but are not central to the Greek verbal system: the participle realizes the factive presupposition, and the infinitive does not realize this semantic feature (see S. Porter 1989: 107, 390–91). In that sense, they are often treated in relationship to the nonassertive attitude.

The causality system is realized by the Greek voice system, which is also a part of the ideational metafunction and is part of the transitivity system, indicating who does what to whom and in what way. There has been much recent work pushing forward discussion of voice in Greek (based on Halliday 2014: 336–45; see Fletcher 2020, whose analysis I draw on, especially regarding the medium-process core; Mathewson 2021; these scholars are paving new ground in this area). Three issues are often raised in discussion of voice: deponency, the meaning of the middle voice, and the meaning of the passive voice. These are important but relatively narrow issues compared to the larger issue of the entire Greek causality system. The major question is whether the Greek voice system is a two- or three-voice system, realized in a bipartite or tripartite causality system. On the basis of historical development—in which the active and then middle/passive forms (especially seen in the present and perfect tense forms) are thought to have developed first, and then the theta forms (that are found in the aorist tense form and future form)—there are those who posit a two-voice system of active and middle, with passive defined as a contextual subcategory of the middle (e.g., R. Aubrey 2016). This may be the history of development of these formal endings, but by classical times and continuing into the Hellenistic period, Greek had a three-voice system. The observation that the theta endings are primary is not an indication of their relationship to active voice, but a recognition of the limitations of morphology and the need for a paradigmatic approach to morphology.

The three voices have active, middle (or, for the present and perfect tense forms, middle/passive), and passive forms. The question is how that voice system semantically functioned. The active voice realizes an active transitive system of direct agency by the subject through the process either to a recipient or not (transitive or intransitive). The middle and passive voices are realized by what is probably best called an ergative system (anticipated in S. Porter 2015a: 29–30; S. Porter/M. O’Donnell 2001) that focuses on the medium, the entity that is involved in the process, and the process itself (the medium-process core); indirect agency is placed elsewhere within the semantics of the clause. The middle voice realizes internal agency that affects the medium-process core; the passive voice realizes external agency, in which the agency is external to the major components of the clause (subject, predicator, and complement) but may be expressed by means of a causal adjunct. The theta form is therefore passive. The present and perfect middle/passive form is either middle or passive in function, but this function must be distinguished by clausal and contextual features as to whether an instance is middle or passive, within the ergative scheme.

This discussion shows that there is no such thing as deponency, in which there are partially realized voice systems with voice substitution. This notion had already been at least questioned by Moulton (1908: 153) and Robertson (1934: 811–13) as well as S. Porter (1994: 72) and then was strongly opposed by Pennington (2003; 2009), since all the voice forms realize distinguishable semantic values: active, ergative middle, or (ergative) passive.

Metaphor occurs at the word group level and especially at the clause level. There are three major types of metaphor: lexical metaphor, interpersonal grammatical metaphor, and ideational grammatical metaphor (Cirafesi 2012; Fewster 2013a; 2016a; Halliday 2014: 659–731; S. Porter forthcoming: chap. 3; Simon-Vandenbergen et al. 2003; G. Thompson 2004: 219–39; cf. J. Martin 1992: 490–91). I do not identify every possible instance of metaphor in the Pastoral Epistles, but all metaphor is based on incongruity and congruity, that is, when wordings are incongruous and point to metaphorical extension. Lexical metaphor occurs when there is incongruity between lexical meanings or lexical domains, such that one entity is being identified with another with which it is incongruous. In other words, lexical metaphor is concerned with differences in meanings among lexemes. Grammatical metaphor, however, is concerned with differences in wording or expression, while retaining a similar meaning. Interpersonal grammatical metaphor occurs in two forms: in the mood/attitude system, when there is incongruity between the expression, such as a declaration, and the semantics, such as a command (see S. Porter 2016f); or when congruous expressions are restated as verbal projections that expand the potential for meaning, such as when a command is projected from a declaration (I want you to go . . .).8 Finally, ideational grammatical metaphor occurs when processes are changed into entities, so that believing is restated as belief. However, such metaphor, at least in Greek and especially because of the participle, also occurs when an entity becomes a process, so that it has all the characteristics of both a noun and a verb (e.g., belief is restated as believing, with tense/aspect and voice/causality, along with gender, etc.).

Beyond the clause, we move to the clause complex (sentence/s) and then to the text. The clause complex is more than one clause and includes a primary clause and secondary and/or so-called embedded clauses (their relation is called taxis, in which they relate either paratactically or hypotactically; see Halliday 2014: 438). Secondary clauses may, as a rule, be described in the same terms as primary clauses, with some restrictions, such as the probability of the subject being expressed as the initial clausal component in relative clauses and the like (subject + X structure). Conditional structures probabilistically have the secondary, inferential clause precede the primary clause. Embedded secondary clauses are based around participles and infinitives, which have the semantic features of verbs but