Exploring Analogy

[an unpublished paper I wrote back in 1979; a bit technical, on the nature of language]
Introduction Religious language is growing respectable again. It had fallen on hard times after the rise of Empiricism, being relegated to the status of "sophistry and illusion," and the forced segregation of religious language from 'this-worldly' language in Kant fairly banished it to the realm of fairy tales. If this were not enough, the positivist label of "meaningless" appeared on it for some time. Ridiculed, segregated, and stripped of value and diplomatic recognition, religious language was the persona non grata of the linguistic world. But in more recent times its worth has been re-asserted. Although still assigned to a separate "language-game", it is now placed on a par with other language games, such as science--which is quite a gain in a world where the scientist has become the religious authority! It is claimed to be the expression of "ultimate commitment" or "way of ordering experience." It is said to have hermeneutic significance to man's being. In this age, in which the messianic character of science is becoming less credible, religious language is growing in respectability.

But although its worth is again recognized, its meaningfulness is still under severe fire. Indeed, religious language is useful but neither true or false. It expresses commitment, not correspondence. But religious language has always had its supporters. There have always been those who have held to and defended its cognitive status. Augustin, Anselm, Aquinas, Calvin all defended the right of religious language to claims of truth- value. And in the history of the battle, the issue of analogy was, and is, central. Macquarrie begins a discussion of analogy noting this fact:

The problem of analogy is a very old one in theological discussion, but it seems to me that it lies very near the centre of the current debates, not only about language but about God and the meaning of the basic Christian doctrines. It is probably the case that many of those engaged in these debates do not explicitly recognize the relevance of the problem of analogy to their work, and may even think that it belongs to an older way of doing theology. In any case, the tendency nowadays is to talk of 'models' rather than 'analogues'--at least amoung British theologians, perhaps influenced by Ian Ramsey. But whatever terminology may be employed, we seem driven to something very like the problem that has been traditionally considered under the heading of analogy. (1)
It is interesting to note that in the above text MacQuarrie virtually equates 'model-talk' and 'analogue-talk.' This relationship is important to note in that the trilemma of equivocity, univocity, or analogy is not transcended by any model-talk that aims at cognivity. Ian G. Barbour is even more explicit in this relationship as he gives a definition of model: "A theoretical model, then, is an imagined mechanism or process, postulated by analogy with familiar mechanisms or processes and used to construct a theory to correlate a set of observations." (2) Because of this relationship, a defense of the validity of model-language of God (or science for that matter!) rests ultimately on a defense of the validity of analogical language of God (or science). But this is often simply a defense of analogy in general, over against claims of univocity of 'other' language realms.

In Defense of Analogy

With the critical importance of analogy for religious language in mind, a defense of analogy in all language may prove a sufficient ploy to cause hesitation among the skeptics long enough to allow us to pursue an investigation of the basis of analogy in theological language. If analogy should prove essential to most (if not all) areas of discourse, we can question the respective areas as to their basis for analogy, provoking them to study while we investigate our own area. This we will attempt.

Its necessity for theology can easily be shown, but this is of no weight to the critics, of course. Macquarrie has a pointed statement of its importance to theology (including non-evangelical theology): "This means that unless we can produce some reasonable account of the logic of analogy, there is no support for our other ways of talking of God, except the via negativa; and, taken in isolation, this leads straight to atheism." (3)

Science fares little better than theology does in this, as is apparent from the dependence of scientific discovery and explanation on theoretical models. Even though philosophers of science are divided on the nature of scientific models, all are agreed that they play a central role in the scientific enterprise. (4) And given the relation of model and analogy accepted above, science is seen to be also dependent on analogy in language and thought.

Theology and science are joined by philosophy at either the intellectual gallows or at the Royal Academy of Science, depending upon the reliability of analogy. Nothing could be more obvious in either the so-called non-metaphysical philosophies which focus on 'seeing as' (an explicit statement of simile !!!) or the deliberately- metaphysical thinkers. Obitts notes that

For those thinkers with metaphysical sensitivities, the 'seeing-as' approach typically develops into the view that metaphysical theories are analogies.(5)
He goes on to quote Dorothy Emment:
As analogies of being, their metaphysical theories seek to say something about 'reality' transcending experience, in terms of relations found within experience. As co-ordinating analogies, they seek to relate diverse types of experience by extension of a key idea derived from some predominant intellectual or spiritual relation.(6)
This dependence on analogy seems strangely unsuited to a discipline which claims to be so rigorous. Yet the dependence is there--acknowledged or not. In a recent article exploring this relation Paul de Man of Yale has expressed this question of dependency as a dilemma:
"It appears that philosophy either has to give up its own constitutive claim to rigor in order to come to terms with the figurality of its language or that it has to free itself from figuration altogether. And if the latter is considered impossible, philosophy could at least learn to control figuration by keeping it, so to speak, in its place, by delimiting the boundaries of its influence and thus restricting the epistemological damage that it may cause."(7)
The writer examines key epistemological texts from Locke, Condillac, and Kant. Finding even the distinction between the figural and the "proper" to be expressed and discussed in these philosophers in figurative terms, de Man arrives at the conclusion that it "turns out to be impossible to maintain a clear line of distinction between rhetoric, abstraction, symbol, and all other forms of language.(8) Of striking significance is his lucid conclusion at the end of the discussion of Kant's Critique of Judgment (section 59):
"If the distinction between a priori and symbolic judgments can only be stated by means of metaphors that are them- selves symbols, then Locke's and Condillac's difficulties have not been overcome. Not only our knowledge of God, to which the passage under examination returns at the end, but the knowledge of knowledge is then bound to remain symbolic." (9)
Having argued that theology, science, and philosophy are all critically dependent on analogy, let me merely note other aspects of our cognitive life that are bound up with analogy.

Quine claims that all language acquisition is dependent on recognizing analogous contexts for use of a word.(10) Mascall points out that the application of the transcendental terms (i.e. being, truth, good, life, etc.) to anything requires analogy.(11) Burrell, in an excellent treatment of the subject, points out that in actuality there is no adequate division between univocal and analogous.(12) This last point comes close to saying that all language is analogical. Indeed, more and more research is pointing in that direction.(13) A quote by Cassirer is representative of much linguistic study on the issue:

"But if this is indeed the case--if metaphor, taken in this general sense, is not just a certain development of speech, but must be regarded as one of its essential conditions--then any effort to understand its function leads us back, once more, to the fundamental form of verbal conceiving." (14)
Some Observations

  1. If analogy is an inescapable literary feature of language, then it becomes epistemologically necessary. In the Biblical worldview this necessity is correlated with man's ontological structure. Hence, we are built to function in this way--another aspect of our ontological status as derivative persons (imago dei).

  2. The very concept of univocity in a referential act can be seen as a special case of analogy. This was, in substance, the remark of Quine above. In the case of naming the same type of object on two separate occasions, the objects are not identical, but are still properly called the term. The same term (formally) is applied to two different contexts If it be objected that this is not a change in the modus significandi, the objector would be hard pressed to even explain this charge. Mascall has explored this in his famous example of 'cabbage- life' and 'human-life'.(15) In applying the term 'life' to both cabbages and men, who could univocally delineate what the difference is, in the modi of cabbage and men? The only way to distinguish the modi significandi is by using the phrases "according to cabbage-ness" and "according to human-ness." Mascall recognizes the problem as one of infinite regress and tries to solve it by balancing an analogy of attribution with an analogy of proportion. This solution has not reduced the concept any further, of course.

    What emerges from this argumentative path, is that all predication using a generic term is analogy.(16) If this is granted, then the univocal cases of predication (N.B. not definition) are only 'apparently univocal' cases. And once analogy is allowed to stand as epistemologically legitimate--by default-- then no point in heaven or earth is safe from analogical description.

  3. Epistemological certainty might be understood in this reference frame. The model concept in science is often associated with 'epistemological immediacy' and vividness. Ramsey's phrase "when the penny drops" points to the self- authenticating character of mediated concepts. Although development of a full-blown epistemology is far beyond the scope of this paper (and probably of its author as well), this area of discussion could provide fertile ideas for such a development. Such a system would certainly contain the following basics:

    Whether or not this proposal could stand bears little relation to the point that generated the above propositions. The point was that if certainty could be mediated through analogy, then the notion of a certain knowledge of God, cognitively mediated through language might be possible.

  4. A simple but significant observation is that the validity of analogical predication is undeniable. If all language using generics (whether explicit, semantic ones such as nouns, or implicit, syntactical ones such as grammatical conventions) is analogical, then any statement denying analogy would be self-stultifying. So, a skeptic can no longer legitimately deny a knowledge of God because of the analogical character of God-talk, Rather, he is now forced to object that the analogy we use is an improper one (implying that there may be a proper one somewhere and the knowledge required to judge propriety!).

  5. This understanding of analogy as composed of a generic, transcategorical, univocal element and an undefinable irreducible modus significandi, also provides a base from which to understand the common epistemic basis of all disciplines. Analogy allows a statement such as "that statement is true" to apply to all "realms" of statements such as ethics, meta-physics, science. And coupled with the linguistic tool of transformational grammar, we see analogy as providing a base context for even discussing other more specific "contexts of meaning" (Nygren) and a language-game in which to discuss specific other "language-games" (Wittgenstein). Each area of more specificity would be descriptive of a set of analogical statements with similar modi.

  6. It is important to realize that all future penetration into the subject of analogy will presuppose the validity of analogy as a truth-vehicle. This is similar (if not integral) to the facts that any discussion of epistemology presupposes its validity and that any discussion of the ability of language to convey correspondent truth presupposes that.

The Character of Analogy

(In the light of our above observation, it is interesting to note that the inescapability of analogy is demonstrated in the title, or any paraphrase thereof, of this section.)

The essential elements in an analogy are the univocal definition (generic) and the specific application. These elements can be seen in relation in Geisler's statement:

"For generic concepts are univocal when abstracted but analogical when asserted of different things, as man and dog are equally animal but are not equal animals. That is, "animal" is defined the same way (say, as "a sentient being"), but animality is predicated differently of Fido and of Socrates. Socrates possesses animality in a higher sense than Fido does.(17)
In this statement we see that all predication involving generics involves analogy, for all generics are, by definition, univocal. When the concept of the generic is correlated with the theory of verbal types, there does not appear to be any a priori reason to distinguish between finite and infinite as belonging to the class. (18) The class is defined only by the decisive traits, not how they are "possessed" by those objects. And if someone argues that the "how" can be placed into the generic term, then (1) a higher level generic term can be found in which that particularity can be transcended (bringing the problem up again) and (2) we can press the arguer to specify univocally the content of the "how" (without generics!).

If, then, analogy does not preclude the class of infinite typological "targets" by definition or nature, then infinite- talk is not really "extension of finite language to infinite concepts." Both finite and infinite analogical language may stand on the same ground of legitimacy. In fact, it can be forcefully argued that the finite references are parasitic on the infinite references. St. Augustin was one of the first to argue this and Aquinas defined his ascription of simple perfections to God in similar fashion.(19) Packer assets this point by arguing that such "extension" is really the function of language (and hence not extension proper):

By depicting God as the first language user, Genesis shows us that human thought and speech have their counterparts and archetypes in Him. By telling us of Adam, Eve, and their descendents listening and responding to God, Genesis shows us that references to the Creator do not 'stretch' ordinary language in an unnatural way; rather such 'stretching' is actually language's primary use.(20)
(Notice that Barth's basic disavowal of analogy as being an extension of the human into the divine falls to the ground within this context. (21))

The Basis for Analogy

The basis of analogy is well-known to be similarity or likeness. (MacQuarrie's view of affinity is not strong enough to provide the univocal basis of the generic.(22)) But even the basis of analogy cannot be discussed without the use of analogy. For example, in the concept of "likeness" or "similarity" the concept is one of "possession" of identical "properties". So it may be that in investigation of the basis we may end in only uncovering another necessary presupposition to language (i.e. similarity as basis for analogy).

Now the notion of "similarity" is surprisingly deceptive, for as often as you attempt to define it, you invariably come up with something like "the basis for analogous predication"! Imagine the following dialogue:

The point is that similarity presupposes analogy and vice versa.

If we know that we can speak analogically of x and y, then we can posit a similarity, even if we cannot specify or conceptualize the point of identity beyond what is already stated in the predications. To affirm 'similarity' is thus only to affirm an analogy. But notice this important fact: similarity is entirely independent of the other properties of the relata. This means that the modus significandi (corresponding to the modus essendi) in an analogy does not affect the ontological similarity--the modi are not considered in the univocal element of the generic.

Similarity of God and the World

The first thing to note here is an observation from the above conclusion. Similarity can exist regardless of the diversity of the modus essendi. That means that God can be similar to the world even though He is infinite and archetypal and the world finite and ektypal.

With this in mind, the acute necessity of similarity as expressed by Lyttkens need not cause alarm:

"The theological significance of analogy may be summarized by saying that some connexion between God and the world is essential. We are then speaking of the analogy based on and implying a real likeness between God and creation.(23)
There have been throughout the history of Christian thought many attempts to explicate the similarity of God and the world. Lyttkens does a good job in tracing changing conceptions of this from Augustin through Anselm, Alexander of Hales, Bonaventura, Albertus Magnus and finally St. Thomas. He points out that Augustin based his analogy (i.e. affirmation of similarity) on the concept of the creature's participatio in God.(24) Anselm was the first to use analogy as a formal category and Alexander of Hales tied participatio to the cause-effect relation.(25) Bonaventura was the first to distinguish Modus from res but it was Albertus Magnus (St. Thomas' mentor) who taught that 'all that exists in anything does so in accordance with its special mode (modus) of existence.' (26).

When we arrive at St. Thomas, we see him using several types of analogy and basing the similarity squarely upon the fact that the effects pre-exist in the cause.(27) This doctrine has come under attack often, forcing a careful statement of the position. A philosopher sympathetic to Thomist thought has offered this careful and precise statement:

"In summation, the analogy between creature and Creator based on causality is secured only because God is the principal, intrinsic, essential, efficient cause of the being and perfections of the world. In any other kind of causal relationship an analogical similarity would not necessarily follow. But in an analogy of being similarity must follow, for being communicates only being, and perfections or kinds of being do not arise from an imperfect being. Existence produces only after its kind, viz. other existences. (28)
In this scheme, the transcendentals (ens, res, unum, aliquid, verum, bonum) are not only in God but are God. That is, God's essence is the standard or reference point for being, truth, life, etc. Personal predicates (wise, holy-, etc.) are likewise intrinsic to God, but not just in that God is wise or holy but that His character is wisdom or holiness. As ontologically ultimate, He is the epistemologically univocal. Thus the perfections of the creature find their standard and definition in the character of God--a univocal point. But the efficient (as opposed to material) causality of God creates the differences in the modi significandi.

Whereas Thomist thought related everything back to God as Cause, later Reformed Scholastics related everything back to God as Intellect. We find statements like this: "The divine ideas of the things created are forms existing in the divine mind from eternity, not really distinct from the divine essence, but which are actually the same as the divine essence." (29) While this comes dangerously close to making the forms eternal and necessary, still this does provide formal similarity of creature and Creator. (How the problems are handled would be of critical importance, of course.)

This idea, if placed into a Thomist frame of God's self- knowledge, would imply that God's knowledge of Himself (and hence of the world as pre-existing in the Cause) produced forms in God's intellect that were structurally identical with created reality. How could this be?

If God's intellect knows the effects (including their form) because He knows Himself as their cause, then this knowledge of the creation is included in but is not identical to, God's self-knowledge. In other words God's essence includes the potential (not potency) for ad extra works--potential for creating (i.e. hypostasizing or reifying) relations. Since the creature cannot be related to God's essence (or it would be necessarily existent and hence God), it is related to God's intellect/will as Designer/Chooser from among other possible worlds. With an infinite number of possible worlds, God's essence then includes the forms of all possible worlds. Thus God's essence can be seen as including infinite potential for ad extra works (to be distinguished from Thomist potency/act categories) and hence, self-knowledge would include the forms (identical with essence) of this created reality. The similarity is shown to reside in the knowledge which God has of Himself--the ultimate reference point of all predication.

Although there may be better ways to configure this situation, the point seems clear. The Thomist cause/effect scheme provides for similarity between the essence of God and the creature; the relation between God's intellect and nature provides the univocal point for analogy. And the creation of derivative subjectivities and objectivities after the pattern existing in the divine essence and actualized by the decision of the Intellect, created the access path of analogy by creation of the first different modus essendi.


All language and knowledge is analogical. We are analogical beings, ontologically and epistemologically, created by a God who 'theomorphized'. Skeptics who would repudiate religious language as being 'only analogical' must now try another tack. They- too use analogy in every generic statement and to provide an ontic basis for this is very difficult in the skeptic's anti-theist system! This relegation of all language to analogy is not loss but gain to the believer, for although it might seem to undermine some univocal statements, it rather guarantees a univocal element in all discourse. A special language of God is not required.

Similarity is seen to be the basis of analogy and only univocal definition can orient us to the content of the identity. The similarity of God to the world can be seen in different perspectives, with God as Cause and Intellect providing an adequate basis for analogical religious language.

The believer need not wear the 'persecuted minority' group feeling. Both he and his language of God fit in an analogical universe.


1. John MacQuarrie, God-Talk: An Examination of the Language and Logic of Theology (London: SCM Press Ltd, 1967), reprinted by Seabury Press in 1979, page 212.

2. Ian G. Barbour, Myths, Models, and Paradigms, (New York: Harper and Row, 1974),p. 30.

3. MacQuarrie, pp. 214-5.

4. Barbour, pp.29-48.

5. Stanly Obitts, "The Meaning and Use of Religious Language" in Gundry and Johnson, Tensions in Contemporary Theology (Chicago: Moody Press, 1976), p.133.

6. Dorothy Emment, The Nature of Metaphysical Thinking (London: Macmillan, 1957),page 215, cited in Gundry and Johnson, Tensions, p.133.

7. Paul de Man, "The Epistemology of Metaphor" in Sheldon Sacks' On Metaphor (Chicago: Univ. of Chicago Press, 1978), p.11.

8. de Man, p.26.

9. ibid.

10. William van Orman Quine, "A Postscript on Metaphor" in Sacks, On Metaphor, p.160.

11. Eric Mascall, "The Doctrine of Analogy" in Religious Language and the Problem of Religious Knowledge, edited by Ronald E. Santoni (Bloomington,Ind.: Indiana University press, 1968), pp. 157-158.

12. David Burrell, Analogy and Philosophical Language (New Haven: Yale University Press, 1973) ,p.220.

13. see the entire volume referenced above in note 7.

14. Ernst Cassirer in Language and Myth, cited in the forward to Sacks On Metaphor.

15, Mascall, pp. 161ff.

16. It could even be argued (if one were willing to tackle Heraclitus) that proper names also were analogous when applied to the same being who was never exactly the same twice. Defenders of univocal predication would have to answer Heraclitus first.

17. Norman Geisler, Philosophy of Religion, (Grand Rapids: Zondervan, 1974),p.281.

18. E.D.Hirsh, Validity in Interpretation (New Haven: Yale University Press, 1967), pp. 49ff.

19. Hampus Lyttkens, The Analogy between God and the World: An Investigation of its Background and Interpretation of its Use by Thomas of Aquino (Uppsala: Almquist and Wiksells Boktwyckeri AB, 1952) ,p.120; Battista Mondin, The Principle of Analogy in Protestant and Catholic Theology (The Hague: Nijhoff, 1963) pp.95-96.

20. J.I. Packer, "The Adequacy of Human Language" in Geisler's (ed.) Inerrancy (Grand Rapids: Zondervan, 1979), p.214.

21. This objection is summarized in MacQuarrie, P.48.

22. MacQuarrie, P.220.

23. Lyttkens, p.477.

24. ibid, pp.113-120.

25. ibid, pp.122-124.

26. ibid, pp.150,154.

27. Mondin, pp.85-87.

28. Geisler, Philosophy of Religion, P.285.

29. Heinrich Heppe, Reformed Dogmatics (Grand Rapids: Baker Book House, 1978) ,p.192.

The Christian ThinkTank...[http://www.Christian-thinktank.com] (Reference Abbreviations)