Copyright 2022 by Randy Leedy, all rights reserved

In Part One of this study I stated that John 21 is the only passage in John’s gospel and epistles that combines the use of ἀγαπάω and φιλέω so extensively. Let’s look more closely at some statistical details. For now, we will restrict our attention to these specific verbs, without factoring in any of the cognates, such as the very common noun ἀγάπη. John uses ἀγαπάω a total of 68 times: 37 times in the gospel, 28 times in the first epistle, twice in the second, and once in the third. He uses φιλέω 13 times, all in the gospel. These counts are the same in major editions of the Greek NT representing the various textual philosophies: Nestle-Aland 28th edition, the Robinson-Pierpont Majority Text, and Scrivener’s Received Text.

Restricting our investigation to John’s gospel for the moment, we find that 74% of John’s “love” verbs (37 of 50) are ἀγαπάω. The passage in chapter 21 that we are considering uses ἀγαπάω twice and φιλέω 5 times. These counts yield a percentage of 71.4% φιλέω, almost exactly reversing the proportions of the overall usage. Since the 5 uses of φιλέω in this passage account for nearly one-third of John’s overall use of that verb, it follows that the dominance of ἀγαπάω outside chapter 21 must be significantly greater than the overall 74%. So let’s calculate that percentage. The 35 occurrences of ἀγαπάω against 8 occurrences of φιλέω outside chapter 21 yield an 81.4% majority use of ἀγαπάω. So φιλέω constitutes only 18.6% of John’s “love” verbs in the first twenty chapters of the gospel, but within our passage, it constitutes 71.4%. In chapter 21, then, John chooses φιλέω over ἀγαπάω 3.8 times as often as he did elsewhere in the gospel.

This difference in relative frequency of the two verbs in these two portions of the book calls for a careful inquiry into possible reasons for such relatively frequent preference for φιλέω in this passage. Perhaps evidence will show that this is nothing more than random distribution, much like highway traffic often flows in clusters, even in the absence of any impediment to normal flow. Or perhaps we will conclude otherwise. It is far too early to jump to conclusions. Even the simple task of collecting word-count information involves some complexities that we have not yet considered. So let’s proceed along that line next, as the main task of this installment of our discussion.

The disproportionate use of φιλέω in John 21 is amplified if we expand our dataset to include the epistles. As noted above, the occurrences of φιλέω in the epistles number precisely zero, a point whose significance will receive consideration later. Recalculating the percentages to include the epistles lowers the proportion of the use of φιλέω outside John 21 from 18.6% to 11.1%, and the factor by which the proportion of this verb’s use in John 21 differs from that elsewhere rises from 3.8 times to 6.4.

One question to examine is whether the 8 occurrences of φιλέω outside John 21 are concentrated in a few passages or whether they are more scattered. If John 21 is not the only passage that features a concentration of these words, then perhaps what we are observing is simply the kind of clustering that randomness often produces.

Only one passage other than John 21 uses φιλέω more than once in close proximity: the word appears twice in John 16:27 (“For the Father himself loves you, because you have loved me and have believed that I came from God” [ESV]). Ἀγαπάω does not appear in this passage, so this is not an instance of alternation between the words like we find in John 21. That fact is perhaps significant. John does not hesitate to use ἀγαπάω repeatedly without variation; here he uses his less frequent verb the same way. An impression begins to emerge that John’s choice between these two words may well be constrained by the meaning he intends to convey, since in many passages, the meaning involved will naturally remain stable throughout the discussion. On this line of thinking, the uniqueness of usage in John 21 would be easily explained: that passage is the only one that explores different degrees or kinds of love. But we are only remaining alert to possibilities right now; we are not yet forming conclusions.

Another passage uses φιλέω twice but at greater distance: the account of the raising of Lazarus in John 11 (vv. 3, 36). This passage more than any other resembles John 21 by also including ἀγαπάω once (v. 5), and the two occurrences of φιλέω are juxtaposed more closely in narrative function than the space between them might lead us to expect, so this passage will require close attention in due time.

Another kind of question affecting our statistical work arises: should our dataset include the respective cognate nouns ἀγάπη and φιλία? Including those words causes the disproportionate use of φιλέω in John 21 to grow by another large step. φιλία appears only once in the New Testament, in James 4:4 (“Do you not know that friendship with the world is enmity with God?”). So the occurrence of φιλέω-family words in John’s writings does not increase at all by including the noun, while including ἀγάπη expands the occurrence of ἀγαπάω-family words by 28 (17 of which are in the first epistle), to a total of 96. The maximal numerical disproportion in the use of φιλέω in John 21 comes by including the nouns and using both the gospel and the epistles as our database: the proportion of “love” words in John 21 that are φιλέω (71.4%) is 9.1 times higher than in the remainder of the gospel and the epistles, which now falls to just 7.8%.

One might decide not to stop even here: these two word families also include the adjectives ἀγαπητός and φίλος, which occur, respectively, 10 and 8 times in John’s gospel and epistles. To include these words would therefore considerably reduce the disproportion of the occurrence of φιλέω in John 21, since the frequency of φιλέω words outside John 21 would double. The usage of these words, though, does not seem to allow the same potential interchangeability that we observe with the verbs. John uses ἀγαπητός almost exclusively in the vocative, as a term of endearment (“beloved”) with which to address the readers of his epistles. It does not appear at all in the gospel. φίλος appears largely in the gospel, and never in the vocative as a term of address. So the author’s choice between these two potential synonyms seems more constrained, and these words seem best excluded from our database.

One passage in which the choice of φίλος might be significant is John 11, where in verse 11 Jesus says, “Our friend Lazarus has fallen asleep.” We can give that word its due attention when we examine that passage, without including it and ἀγαπητός in the body of statistics. 

If someone wanted to reverse this judgment call, though, and include these adjectives in the counts, then the disproportion of the use of φιλέω in John 21 would fall from 9.1 to 5.4 times the rate elsewhere. While this reduction is significant, the degree of disproportion remains strong.

Once we have decided to exclude the adjectives as not seeming to be candidates for free variation, we must also reconsider, on the same basis, our inclusion of the nouns. The very lopsidedness of the overall NT usage of the nouns, 116:1 in favor of ἀγάπη over φιλία, suggests that some factor is preventing the use of φιλία as a viable synonym for ἀγάπη. For our bottom line regarding the disproportion of the usage of φιλέω in John 21, then, it seems best to go back to including only the two verbs, which yields this statistical statement: the proportion of φιλέω to ἀγαπάω in John 21 is 6.4 times that of the remainder of John’s gospel and epistles. Or, if we decide not to include the epistles, the multiple of disproportion falls back to the 3.8 times that we calculated at first.

Regardless of which dataset one decides to work with, the degree of disproportionate use of φιλέω in John 21 is great enough to create a reasonable burden of proof for the view that John uses these synonyms interchangeably as a simple matter of literary style and not because of any distinctive meaning that either conveys. That said, it is also reasonable for my own position to assume a burden of proof as well, since, random clustering is always available as one possible explanation. In the end, one can compare the support offered for the competing positions and judge which is most convincing.

We will now begin to consider some factors that diminish the likelihood of random clustering as the best explanation for the disproportionate use of φιλέω in John 21. The reader may already have noticed and perceived the significance of the fact, mentioned above, that the whole body of usage of the “love” verbs in the epistles (31 occurrences) is 100% ἀγαπάω. The gospel, too, contains a major passage dominated by ἀγαπάω. That verb occurs 21 times in chapters 13-15 (accompanied by 5 occurrences of ἀγάπη), which is well over half the gospel’s occurrences of each of those words. Never once is that series interrupted by an occurrence of φιλέω. The cognate φίλος does appear, but its sense, “friend,” as opposed to “slave,” does not have a counterpart cognate of ἀγαπάω that could have been used instead. Only at the very end of that series, in a statement where a distinction in meaning can easily be made out, does φιλέω appear: “If you were of the world, the world would love you as its own” (15:19).

Chapter 11 is the only passage other than chapter 21 that combines the two verbs with any alternation: φιλέω appears twice (vv. 3, 36) and ἀγαπάω once (v. 5). Compared to John 21, this passage contains fewer uses of these verbs spread over more verses, so it is not a strong parallel to provide a second example in support of free variation in John 21. Furthermore, as in the case of φιλέω in 15:19, a distinction in meaning can easily be supported. [Note: extending this passage from vv. 3, 5 to include v. 36 weakens a claim that I made in Part One, where, excluding v. 36 due to its distance, I said that no other passage uses these verbs with anything remotely approaching the combination employed in John 21. I have now revised that section of Part One to soften that language.]

A third passage containing 100% ἀγαπάω for its “love” verbs is 17:23-26. In this passage the verb occurs 4 times, with no occurrences of φιλέω in the entire chapter. 

Whatever might be said about the significance of these numerical observations, one thing is abundantly clear: John does not hesitate to use one verb or the other repeatedly, without stylistic variation. A claim that that the variation between ἀγαπάω and φιλέω in John 21 reflects John’s regular stylistic pattern of deploying these two verbs is entirely untenable. The possibility that John uses the two verbs synonymously in separate passages remains open, but this is hardly a matter of literary style. The stylistic issue is the avoidance of redundancy, which is relevant only to repetition within a brief segment of text. Someone who wishes to hold that the variation in this passage is merely stylistic will need to advance a plausible explanation for why John has chosen such a style for this passage alone—or at least, to give some weight to our findings in chapter 11, practically alone.

Another important fact to ponder is that reversal in which of these verbs dominates John 21 in contrast to the remainder of John’s gospel and epistles. One would think that if John’s only purpose in varying the word choice in chapter 21 is merely stylistic, surely he would maintain his general preponderance of ἀγαπάω and just sprinkle in a few occurrences of φιλέω. What motivation other than communicating distinction in meaning could possibly account for the reversal? I ask this question, not rhetorically, as equivalent to asserting an impossibility, but as a serious question to be entertained as this study progresses through further installments.

An important question may have occurred to many readers: Why have I excluded Revelation from the database of word occurrences? Do I doubt its authorship? No; the reason is simply the fact that the “love” verbs are relatively infrequent in Revelation. They appear 6 times: ἀγαπάω 4 times and φιλέω 2. Worth noting is the fact that Jesus uses both verbs in chapter 3 for his love for his people, much like the gospel uses both verbs for the love between the Father and the Son (an observation that, by the way, provides at least very minor support for traditional Johannine authorship of Revelation). Since such usage is already present within the gospel, Revelation contributes little to this study, so I have opted for the smaller body of literature as the study’s primary database.

To move toward the conclusion of the numerical aspect of this study, let’s quantify the relative usage of these verbs in the portion of the New Testament that so far we have excluded: the non-Johannine books and Revelation. While the usage within the body of John’s writings is most relevant to our study in John 21, any general patterns of usage that we might find elsewhere in the New Testament can potentially provide further light.

In the books we are now exploring, ἀγαπάω occurs 75 times in the Nestle-Aland text (74 times in the Majority and Received texts, where Jude 1 reads “sanctified” rather than “[be]loved”). Φιλέω occurs 12 times in all these editions, 3 of which are in the sense “kiss” (all referring to Judas’ betrayal of Jesus) and therefore are not instances of a “love” verb. So the percentage of “love” verbs that are φιλέω in these books is 9/84 or 10.7%. This is remarkably close to the 11.1% that we observed across John’s gospel and epistles excluding John 21. Other than the account we are studying, then, John’s overall proportion of relative usage of these two verbs is right in line with the rest of the New Testament. There is basically no clustering of the 9 occurrences of φιλέω outside this Johannine material; the only passage using the verb more than once is a poetically parallel statement in Matthew 10:37, “Whoever loves father or mother more than me is not worthy of me, and whoever loves son or daughter more than me is not worthy of me.” Nor do any of these 9 occurrences combine with neighboring occurrences of ἀγαπάω. The repeated and combined use of these two verbs in John 21, then, is unique not only within John’s writings but also within the whole New Testament.

Should we expand further still and include the Septuagint? Φιλέω occurs there (Rahlfs’ edition, limited to the canonical books) 18 times in the sense “love.” (The sense “kiss” appears 10 times.) This is against 211 occurrences of ἀγαπάω, for a percentage of just 7.9%. The only clustering, 3 occurrences, is in Genesis 27, where, before Isaac blesses Esau (or such, at least, was his intent!) he wants the food that he loves. Some interesting juxtaposition of the two verbs will require attention within the word study yet to be developed, but nothing is extensive enough to rival John 21 or to weaken the claims that I intend to demonstrate in favor of a distinction in meaning between these verbs in that passage. In short, a cursory overview of Greek Old Testament usage casts no significantly different light upon the previous numerical findings.

Let’s return now to the matter of John’s pattern—other than in chapter 21 and to a lesser extent in chapter 11—of using the same verb for love repeatedly, without variation, within contexts of various lengths. Given that reality, the exegete who is skeptical about claims of mere stylistic variation between these two verbs in John 21 hardly deserves to be dismissed as some poor soul destitute of the wisdom to be found in modern Linguistics. I do not mean this statement to deny the value of Linguistics. I am only saying that a superficial application of Linguistics does not negate the findings of careful exegesis. Lest I be misunderstood, though, I must hasten to add that what we have undertaken so far, which is not much more than a simple comparison of word counts, hardly qualifies as careful exegesis. So the reader will be patient, I hope, in awaiting more analysis to come. I have no fear that that analysis, when fully presented, will lightly be dismissed as lacking thoroughness.

Just to foreshadow good things to come, let’s go, in John 21, one step beyond the mere word counts. Apparent right on the surface is a suggestion of pattern rather than randomness in the variation between the two synonyms The two occurrences of ἀγαπάω both occur in the mouth of Jesus, in his first two questions to Peter, who, throughout the conversation, consistently replies with φιλέω. Then the final four instances of “love” verbs are all φιλέω, as though the focus of the conversation moves from one verb to the other. Simple variation certainly could produce this arrangement, but a writer mixing his vocabulary synonymously for the sake of style might well be expected to help his reader recognize that intent by employing a greater mixture of placement of the two words in their order and in the mouths of the two speakers.

This impression of a pattern somewhat intensifies the hunch, based on the relative word frequencies, that meaning-motivated intentionality may well be at work here. And the impression of intentional pattern gains strength from John’s handling of the synonyms for feeding and for sheep throughout the passage. That handling seems to signal, by lack of any discernible patterning, that these pairs of synonyms are to be understood as functionally interchangeable—the differences in denotation, which are unquestionable, do not affect the basic communicative intent of the statements employing them. “Feed my lambs” and “Shepherd my sheep” both communicate Peter’s call to pastoral ministry. This presence of other synonym pairs within the conversation has often been used as evidence that the “love” verbs are also used interchangeably. But if closer examination finds that the various pairs of synonyms are mixed differently enough to create a contrast between their handling, then the presence of those additional synonym pairs may well argue for the opposite conclusion and support our confidence that John’s arrangement of the “love” verbs is intentional and significant for meaning. This point will receive more detailed treatment in a later installment of this study.

The next installment of this study will take up the question of what distinction in meaning between ἀγαπάω and φιλέω might prove to be present in John 21.