This summer my fiancé and I have been preparing for our wedding, which will be in September. One aspect of our preparation has been reading through the workbook Before you Say “I Do”: A Marriage Preparation Manual for Couples. Admittedly, it was first published in the 1970s (reprinted in 1997), but generally I have found the content to be quite good and thought-provoking. However, in the chapter on “Love as a Basis for Marriage,” the authors make a distinction between three kinds of love: eros (ἕρος) (romantic, sensual, or sexual), philia (φιλία) (friendship, companionship, cooperation), and agapē (ἀγάπη) (kindness, personal act of commitment). For those of you who know the biblical terms for love, philia and agapē will be familiar to you. I have seen (and heard) this kind of distinction made before in books, sermons, and Sunday school classes. It is a popular way to talk about love in the Bible and seems to make sense. Yet I believe it is fundamentally mistaken.
Lessons in Etymology
When studying words in any language, it is important to note that any one word has a semantic range; this means the word has a range of appropriate definitions depending upon the context in which the word is used. A semantic range can be large or small, narrow or broad. For example, “to save” in Greek (sōzō; σᾡζω) can mean 1) to save from physical danger, or 2) to save from spiritual death (BDAG, p. 982). These are quite different, yet the same word can be used for both meanings. In fact, sōzō can be used in both ways in the same pericope (e.g. James 5:13-20). How do you tell when each meaning is being used? Context. Context is always the number one determinative factor in translation issues like this (lower levels of validation include the author’s style, grammar, and biblical theology). In addition to this, each word has its own unique semantic range. Although we talk about words being synonyms, there are no true, or perfect, synonyms. In reality, synonyms are words whose semantic ranges overlap to a great extent; yet there will always be some slight difference between them. Still, there can be instances where you will have words that are used snonymously in certain contexts because they are pointing to the same referent, even though by themselves (and divorced from context) they do not mean the exact same thing over their total semantic range. 
I am not saying that there aren’t legitimate distinctions between eros, philia, and agapē since there are. However, when translating these words, one must not look primarily to the word’s definition, or to its origins (etymology) to discover their “true” or “real” meaning. To do this would be to commit the root fallacy, which defines the meaning and use of a word by its origin alone. A simple example will suffice to show this fallacy: in English, the word “butterfly” has absolutely nothing to do with its constituent parts, “butter” and “fly,” which when isolated meaning something completely different than when conjoined. Greek has many similar constructions as well. This isn’t to say that sometimes the origin of a word or its parts doesn’t add up to its complete meaning, since for example, ekballō (ἐκβάλλω) is composed of the preposition ek (ἐκ) (out of, from) and the verb ballō (βἀλλω) (to throw) and literally means “I cast out,” “I throw out.” The point is that one should not automatically default to etymology to solve linguistic and definitional issues in Hebrew, Greek, or English.
Before I explode this myth of the three kinds of love, I want to look briefly at the semantic ranges of these words. First, you should be aware that eros never actually occurs in the Bible; although it was used in Koinē Greek, none of the biblical writers, either in the Septuagint or New Testament, ever use this word. In the classical (Attic) Greek lexicon (LSJ), eros is defined as “love, desire.” There is no intrinsic part of the definition of eros that points toward any kind of sexual or romantic love.  In fact, there is good evidence that eros in fact did not have such a restrictive definition. Eros in ancient Greek thought signified a search for wholeness and completeness. This actually makes sense of the sexual and romantic connotations that have come to be associated with the word since by God’s design, sexual intimacy within marriage is designed to bring an interpersonal oneness and wholeness (or in biblical terminology, “the two shall become one flesh”). 
Philia (φιλία) means “friendship, love” (BDAG, p. 1057). Its cognate verb, phileō (φιλέω), has a wider semantic range and can mean anything from “having a special interest in someone or something with focus on close association, have affection for, like, consider someone a friend,” to “to kiss as a special indication of affection” (BDAG, pp. 1056-57). This second definition of “to kiss” signified a later evolution in phileō’s etymology and will come into play later in the relationship between phileō and agapaō (ἀγαπάω; cognate verb of agapē). Αgapē can mean anything from “esteem, regard, love,” to “a fellowship meal, love feast” (BDAG, pp. 6-7). Agapaō’s semantic range includes “to cherish, have affection for, love,” to “to take pleasure in” to “to prove one’s love” (BDAG, pp. 5-6). Just surveying the semantic ranges of these words – their various uses and elasticity depending upon the context – should throw doubt on the linguistic hypothesis that these words have rigid and locked definitions that can be applied mindlessly when encountered in the text.
The Myth Debunked
The particular myth involved in the three kinds of love is just that: that the Bible delineates three distinct kinds of love and the biblical authors knew this so that when they used eros they meant sexual passion and desire, when they used philia they were referring to the affection and bond of friendship, and when they used agapē they were referring to a higher “spiritual” or “God-like” unconditional love of kindness and commitment. To dispel this myth all one would have to do is find examples where these words are used interchangeable to refer to the same kind of love, or where their assumed definitions are clearly violated. If such examples could be found, then this fixed distinction between the three loves will be shown to be invalid.
D.A. Carson, in his excellent little study, Exegetical Fallacies (2nd ed.), addresses this very problem and nicely provides us with these examples:
- In 2 Samuel 13:1, 15 (LXX) Absalom’s desire for and rape of his half-sister Tamar is described using agapaō (v. 1: ἠγἀπησεν, Aorist, 3sg) and agapē (v. 15: τἠν ἀγάπην, accusative, sg).
- In 2 Timothy 4:10 Demas’ desertion of Paul due to his love for the world is described using agapaō (ἁγαπήσας, Aorist participle).
- In the Gospel of John, the evangelist uses agapaō (3:35) and phileō (5:20) interchangeable to denote the Father’s love for the Son. 
- In Matthew 26:48-49 phileō (and kataphileō) are used to describe Judas’ kiss of Jesus on the night of his betrayal. This obviously does not denote true friendship since Judas was using the kiss as a sign to hand Jesus over to be arrested and crucified.
Although examples like this could be multiplied many times over, these four suffice to explode the myth of the three kinds of love. Absalom’s rape of Tamar was an unspeakable evil that brought her and their entire family terrible shame; needless to say, this has nothing to do with God’s unconditional loving-kindness toward us. Undoubtedly agapaō and agapē can both be used to describe godly love but also worldly desire and lusts. Likewise, phileō can be used beyond friendship to refer to a lowly kiss between two people or the most holy and perfect love that is shown between the persons of the Trinity.
Any student of Koinē Greek should be able to recognize these fallacies and (hopefully) not trip and fall into such lexical traps. However, most Christians haven’t had such training, and since this distinction is quite popular in sermons, books, and other literature, it was imperative to spell it out specifically with the hope of helping others avoid such etymological perils.
 I am assuming that eros was used in Koinē Greek; however, when I checked the papyri lexicon by Moulton & Milligan (Vocabulary of the Greek Testament), they did not have an entry on eros.
 For more information on eros see this article (I do not endorse everything in this essay) or listen to this lecture on political philosophy by Yale University professor Stephen B. Smith (between 35:59-36:41). In this case, the context of eros is the quest for the best political knowledge, the highest and most virtuous understanding of how to construct a good and just society.
 Pp. 31-32. Carson also extensively treats the passage in John 21:15-17 where Jesus asks Peter three times if he loves him (using agapaō twice and phileō the last time) and Peter’s affirmative response each time using phileō. Carson clearly shows that many have drawn false theological conclusions due to the contrast between these two verbs and Jesus’ switch to phileō to match Peter’s response in their last exchange (pp. 51-53).