On the Liberal Art of Grammar

by John F. Nieto

Thomas Aquinas College 

1. The following comments propose to clarify the nature of grammar as an art, a speculative and liberal art. First I distinguish grammar from other arts concerned with speech [2-9] with particular attention to the difference between grammar and logic [6-9]. Then I show that while grammar is an art, it is a ‘speculative art’ [10-24]. (Here I show how this art is ‘speculative’ as a whole [10-11], and can yet be divided into parts that are ‘speculative’ and ‘practical’ in several ways [12-25].) Finally, I discuss the respect in which it is entitled ‘liberal’ [26-28].

2. The three parts of the trivium, grammar, rhetoric, and logic, have speech as their subject. Though it is not part of the trivium, one might add poetry to this list. Now, it is not difficult to see that these are distinct arts. Each of them must therefore consider speech in its own way.

3. The division of these arts from one another is made clearer by considering that speech belongs properly to the composite of the human body and the rational soul, which is immaterial. This soul has operations in common with the body, such as fear or anger, as well as other operations proper to itself, such as thinking. Speech can be therefore ordered to something immaterial as well as to something material.

4. Logic considers speech insofar as it manifests some immaterial, universal intention, together with all those things that follow such universality. Both rhetoric and poetry consider speech insofar as it manifests not only thoughts but also the passions common to body and soul. But rhetoric considers speech insofar as thought and passion can be ordered to human action, while poetry considers speech insofar as these passions can be ordered to the pleasure and delight of those listening. Respectively, these three arts consider speech insofar as an honest good, a useful good, and a pleasant good can be found in it.

5. The grammarian, however, does not consider speech precisely as it attains any of these ends. Rather, like many ministerial arts, it considers the making of the instrument as such. A lower art commissioned by a higher art to make its instrument knows the order of this instrument to that end, although it does not know the proper causes of that end. The violin maker knows the order of his instrument to music making, though he does not, precisely as a violin maker, know how to play this instrument.¹Physics 194a33-b7 So the grammarian considers speech as an artifact capable of expressing thought and even passion. (Thus a good grammar has a section on the fundamentals of prosody.) But the grammarian does not consider speech precisely insofar as it attains a further end. Rather, he considers the proper principles by which speech itself is formed. He considers what makes a word to be a noun, a verb, or some other part of speech, and the order these parts of speech have to one another. Thus he ultimately considers the constructions that arise from the order between such words as from their proper causes.

6. Distinguishing grammar more carefully here from logic can assist in seeing the order proper to grammar. Logic considers the order in words precisely insofar as this order manifests the order in thought, which must be resolved to things themselves. Thus, for example, the logician considers ‘substance’ insofar as it is a name signifying some individual, such as Socrates, or its essence insofar as these can be conceived.²Categories 5 Again, the logician recognizes that ‘action’ is represented as belonging to some subject and terminating in an object.³Cf. Physics 202a21-b29, which Aristotle introduces as an ἀπορία λογική. But he considers this ‘mode of signifying’ to be a manifestation of what action is and how it is conceived. Thus, he recognizes that the verb ‘to suffer’ is not in the logical category of action.⁴II Sent. D.35, a. 1, ad 5. For the logician the mode of signifying is always considered insofar as it signifies something with a mode of understanding and thus a mode of being.

7. But the grammarian only considers the order in words insofar as it is a principle of sentences. For the grammarian, ‘substance’ is merely something about which other things can be said. He forms the noun and pronoun with this ‘mode of signification’, whether or not the nature signified is a substance logically: ‘man’, ‘humanity’, ‘whiteness’. Likewise he considers the relation of action to a subject and an object insofar as this produces certain kinds of verbs. Thus he sees that the concept of action has produced a distinct schema or template by which the active, transitive verb is formed as an instrument to his intellect, without attention to the reality signified or its definition. In this way ‘suffer’ in the following passage is understood by the grammarian to be an active, transitive verb: ‘[I]t can be only weak, irresolute characters...who will suffer an unfortunate acquaintance to be an inconvenience, an oppression for ever’ (Emma).⁵Emma c.7 Note that, to my mind other, more fundamental uses of the verb ‘suffer’ bear the ‘middle’ voice.

8. This difference between logic and grammar is pointed out by Saint Thomas when he says,
   quia logica ordinatur ad cognitionem de rebus sumendam, significatio vocum, quae est immediata ipsis conceptionibus intellectus, pertinet ad principalem considerationem ipsius; significatio autem litterarum, tanquam magis remota, non pertinet ad eius considerationem, sed magis ad considerationem grammatici.⁶Expositio libri peryermenias I l.2, p.3: Because logic is ordered to raising knowledge about things, the signification of sounds of voice [vocum], which is immediate to the very conceptions of the intellect, pertains to its principal consideration; but the signification of letters, as if more remote, does not pertain to its consideration, but rather to the grammarian’s consideration.
   Saint Thomas explains the logician’s concern with the signification of vocal sounds by stating that this signification is immediate to the intellect’s conceptions. One could say that the modes of signifying no longer exist in the written word itself to the extent that these modes of signifying concern the logician. For in writing they are separated from thought. Only when the written text is again read can the logician find his object. For his object is never separated from the modes of understanding.
9. But the written word still possesses in some manner the mode of signifying insofar as it is ordered to certain constructions. The subject, object, and verb thus demand certain forms and positions if they will cohere in a sentence, and these properties are found in writing. Thus, one who is learning another language can recognize the grammatical implications of certain ‘cases’ and positions in a sentence without understanding what the sentence says. He notices that canem is accusative and thus some kind of object. Or, again, that ‘man’ and ‘dog’ are in the position appropriate to the subject and object respectively, though he does not know what these words mean. For the modes of signifying are not considered by the grammarian as revealing things and the manner in which those things are conceived, but as constituting parts of speech with the power to be brought together to form a certain whole, the sentence, the mind’s principal instrument for expression.

10. Grammar is in this way an art, that is, a ‘certa ordinatio rationis quomodo per determinata media ad debitum finem actus humani perveniant.’⁷In libros posteriorum analyticorum I l.1: ‘a certain ordination of reason in what way human acts arrive through determinate means to a determinate end.’ It considers the modes of signifying as the means by which one makes speech. And in virtue of its object grammar is a liberal art. For it has ‘opus aliquod quod est immediate ipsius rationis, ut...orationem formare.’⁸Super Boetium De trinitate Q.5, a.1, ad 3, 231-233. Cf. I-II Q.57, a.3, ad 3: some work which belongs immediately to reason, as...to form speech’. The sentences grammar constructs by means of the modes of signifying are immediately instruments of the intellect in expressing its thoughts.

11. Speech, however, is distinguished from the sounds of animals, not by its expression of passion (however much more sublimely it does so), but by its order to the expression and communication of human thought. For this reason the art concerned with speech precisely as a sign, that is, as an instrument of thought, is among the ‘speculative’ or ‘theoretical’ arts which are ordered to particular, yet speculative, ends.⁹Super Boetium De trinitate Q.5, a.1, ad 4, 273-277: Cum enim philosophia vel etiam artes per theoricum et practicum distinguuntur, oportet accipere distinctionem eorum ex fine, ut theoricum dicatur illud, quod ordinatur ad solam cognitionem veritatis, practicum vero, quod ordinatur ad operationem. Hoc tamen interest, cum in hoc dividitur philosophia totalis et artes, quod in divisione philosophiae habetur respectus ad finem beatitudinis, ad quem tota humana vita ordinatur....Cum vero dicuntur artium quaedam esse speculativae, quaedam practicae, habetur respectus ad aliquos speciales fines illarum artium. [For when philosophy or even the arts are distinguished by the theoretical and the practical, one must take their distinction from the end, as that which is ordered only to the knowledge of the truth is the theoretic, but what is ordered to operation is practical. Yet there is this difference, when the whole of philosophy and the arts are divided in this that a reference is had in the division of philosophy to the end of beatitude, to which the whole of human life is ordered....But when certain arts are said to be speculative and certain practical, a reference is had to the some special ends of those arts.] In this sense grammar is always a speculative art, no matter how practical the manner of its study is, no matter how slavish its use is.

12. But, while the grammarian’s consideration of the modes of signifying and the constructions they cause is in itself and as a whole a ‘speculative art’ because it produces an opus belonging immediately to reason, there is reason to distinguish within grammar a part that is speculative from a part that is practical. For the grammarian’s consideration can be ‘propinqua uel remota ab operatione.’¹⁰Super Boetium De trinitate Q.5, a.1, ad 4, 283: ‘near or far from operation.’ Saint Thomas discusses this distinction as it is appropriate to medicine, which as a whole must be judged a practical art in consideration of its end, the healing of the body.

13. For part of medicine too can be called ‘practical’ and another part ‘speculative’. One part of medicine which ‘docet modum operandi ad sanationem’¹¹Super Boetium De trinitate Q.5, a.1, ad 4,: ‘teaches the mode of operating for health.’ can be called practical because it is near the operation considered by medicine. Another part that ‘docet principia, ex quibus homo dirigitur in operatione, sed non proxime’¹²Super Boetium De trinitate Q.5, a.1, ad 4, : ‘teaches the principle from which man is directed in operation, but not proximately.’ is called speculative merely because of its ‘distance’ from operation.

14. A similar distinction between what is near operation and what is far from operation can be found in grammar. And this distinction can be applied to grammar in several ways. Here I will propose three that I understand to be of particular importance.

15. In one way this distinction is found in grammar just as it is found in medicine. This involves a distinction of the consideration of grammatical principles from their application to particular operations. For the grammarian must obviously be able to form particular sentences and correct particular grammatical errors. Teaching how to do so is close to operation and therefore ‘practical’ grammar.

16. But grammar also distinguishes the modes of signifying and recognizes them in the various parts of speech. These modes of signifying are then assigned as the proper causes of the constructions found in speech. In this way a particular category, such as substance,¹³I Sent. D.22, Q.1, a.1, ad 3: Sed grammaticus accipit substantiam quantum ad modum significandi, et similiter qualitatem; et ideo, quia illud quod significatur per nomen significatur ut aliquid subsistens, secundum quod de eo potest aliquid praedicari, quamvis secundum rem non sit subsistens, sicut albedo dicit, quod significat substantiam, ad differentiam verbi, quod non significat ut aliquid subsistens. or even a very determinate nature, such as a ‘chain’ (taken for an unnamed genus),¹⁴Martini de Dacia Modi Significandi c.11. Martin conceives the preposition to arise first for the purpose of completing an intransitive verb in such a manner that it can take an object, as in the sentence, ‘I walked into the building.’ As I read him, he understands the ‘hook’ to exemplify the relation of the preposition to an intransitive verb (at least originally) in the relation of the hook to whatever ‘holds’ the hook. The relation of the preposition to its object (and through it the intransitive verb’s object) is exemplified by the relation of the hook to whatever it ‘hooks.’ is understood to provide the ‘schema’ or mode according to which a particular word is this or that part of speech, here a noun or a conjunction. Again, the nature of the demonstrative pronouns ‘this’ and ‘that’ (and their relation to antecedents) may be considered universally, without considering how to use them in a manner close to speech. The grammarian also shows why some mode of signifying gives rise to some construction: how the mode of action allows the verb to be said of a subject¹⁵Expositio libri peryermenias I, l.5, p.5. or how the mode of a ‘hook’ allows a preposition to terminate an intransitive verb in an object.¹⁶Martini de Dacia Modi Significandi c.12. The grammarian thus considers universally and in principle the operation of all speakers. Insofar as his consideration is distant from the particular act of speaking, it can be called speculative.

17. Note, however, that, though such consideration is ‘remote from operation,’ it is not speculative in the sense that it does not have some opus or is not ordered to operation. It can be distinguished as speculative rather than practical because it is farther from the particular opus and operation than other considerations are. Rather than separating such considerations from operation altogether, this ‘remoteness’ allows the consideration to embrace in a universal manner many more opera and operationes. Thus ‘speculative grammar’ is still an art.

18. Another way in which this distinction is found in grammar arises from the fact that the principles and causes of speech can be considered insofar as they bear upon a particular matter. One often sees, for example, the theologian, who studies the divine nature, consider the application of grammatical principles to the particular matter he speaks of. Sometimes he discusses determinate propositions, as when he explains why the past imperfect is used in the statement, ‘In the beginning was the Word,’¹⁷Cf. I Sent. D.8, Q.2, a.3 ad 5; Super Evangelium Ioannis c.1 l.1. or the sense of the demonstrative pronoun ‘this’ in the sacramental formula ‘This is my body.’¹⁸Cf. IV Sent. D.8, Q.2, a.1, q.4; III Q.78, a.5. Again, the theologian sometimes considers the very possibility of naming God or forming propositions about Him.¹⁹I Q.13, a. 1, a. 1, 12. Now all these considerations are near to operation (though one may be nearer than another) because they are contracted to the matter being spoken of.

19. Sacred Theology must often ask such questions, because its subject matter is not properly represented by the modes of signifying found in speech, but other sciences may do so as well. In natural theology the philosopher recognizes that one must not only say that God is alive, but also that he is his life.²⁰Metaphysics 1072b26-30. Again, the physicist must recognize that the perfect and the imperfect or ‘progressive’ aspects signify something really distinct when said of local motion. ‘To be walking to Athens’ is not ‘to have walked there.’²¹Physics 232a4-6. Yet, when said of the operations of sense, these aspects signify the same reality. ‘To be seeing something’ and ‘to have seen it’ do not signify a different reality. They do signify that reality with a different aspect. One phrase signifies the act as ongoing; the other signifies it as perfect or complete. But, unlike walking, the act of seeing itself is complete (I do not mean ‘over’) as soon as it begins.

20. In these examples, grammar is serving another science. Yet even within grammar itself, especially in the consideration of particular languages, the grammarian may be concerned with the need to speak about a particular matter. Thus he may explain the use of the passive or ‘middle’ voice to name certain actions, such as sensing (αἰσθάνομαι) or following (sequor.) Again, he may distinguish the material and formal accusatives in a statement such as ‘We made him king.’ This distinction is founded on some relation in the things signified, the man and his kingship. Likewise, the distinction in various kinds of genitive constructions (e.g. the possessive, the subjective, or the objective genitive) is founded on distinctions in the matter represented by the genitive.

21. In all such cases, grammatical principles must be applied to the matter at hand, whether this occurs in a particular proposition or a very determinate kind of proposition. This nearness to operation is, it seems to me, one way in which a part of the art of grammar, although it remains part of a ‘speculative art,’ can be called practical grammar rather than speculative grammar. In this determinate sense, grammar would be speculative when it fails to consider the grammatical import proper to the matter spoken of.

22. Grammar can be distinguished as ‘practical’ and ‘speculative’ insofar as its considerations are closer to or more removed from operation in the two ways mentioned. But grammar can also be remote from or near to operation through a cause proper to its subject. For, though the modes of signifying that grammar studies, whether or not found in all languages, have a kind of universality, these modes of signifying only exist in particular languages that embody them in sounds determined by convention and so they vary in one place and another and at one time and another.

23. Thus what is commonly understood as grammar involves the consideration of the determinate words and constructions used by a language or even the comparison of these among various languages. All such considerations are obviously nearer to operation and can thus be called ‘practical.’ But the considerations that abstract from any particular language, even if they illustrate grammar’s teachings with the usage of particular languages, are remote from operation and are in this sense called ‘speculative.’

24. Again, even the determinate considerations mentioned above, by which a science applies the teachings of grammar to a particular matter, may demand consideration of one or more particular languages. So Saint Thomas discusses what is proper to Greek and Latin when commenting on Aristotle’s definition of the verb²²Cf. Expositio libri peryermenias I l.5. or on the prologue to Saint John’s gospel.²³Cf. Super Evangelium Ioannis I l.1, passim. Such considerations would be practical in two ways, insofar as it considers particular sentences and insofar as it is concerned with the peculiarities of one language in distinction from another.

25. Hence, grammar can be divided into speculative and practical parts in at least these three ways, as it is concerned with the principles of speech or their application in forming particular sentences [15-17], or as it considers speech without attention to the matter spoken of or with such attention [18-21], or as it is concerned with the very nature of language or with particular languages [22-24]. Note that in these ways parts of grammar will be called ‘speculative grammar’ or ‘practical grammar.’ But the whole of grammar is not called ‘speculative grammar’ but a ‘speculative art.’

26. Note that all the considerations mentioned belong to grammar insofar as it produces an opus belonging immediately to reason. Such considerations therefore belong to grammar as it is a liberal art. Now an art is liberal insofar as it is ordered to the intellect’s satisfaction.²⁴Metaphysics 982b25-28. Cf. Sententia Libri Politicarum I l.5: dicuntur aliquae artes liberales, quae deputantur ad actus liberorum. In this way ‘liberal’ adds some notion to ‘speculative.’ The speculative art produces some work that belongs immediately to reason, but the liberal art considers that work in a manner that serves man’s intellect and thus makes him free.

27. So any consideration of particular languages ordered merely to obtaining the habits of speaking, reading, or writing that language without attention to the principles by which it is an instrument of the intellect shares little or not at all in the liberal character of this art. Though one cannot make the art ‘servile’ (for one cannot change the nature of the art), one can use the art in a servile way.

28. The liberal character of grammar demands that in the consideration of a particular language, even in its idioms, one sees the order in words as an instrument the intellect forms for the expression of its thought. To the extent that the grammarian fails to consider the order instituted in language by the mind, he fails to understand the order of speech to his own intellect. He thereby does not consider grammar in the manner appropriate to the free man, who lives for his own sake and thus for the sake of the highest faculty.²⁵The grammarian often considers principles outside his science to clarify his own principles. Thus he may, for example, consider the modes of understanding implied in various uses of the genitive: the possessive genitive, the subjective genitive, the objective genitive. This order allows him to understand the nature, unity, and breadth of the genitive case. Again, historical principles may explain the development of the accusative case. Though such principles are outside the science, they help to manifest the nature and unity of this case. A good example of the use of such considerations as an aid to what is properly speculative consideration (however imperfectly it is distinguished from the practical) can be found in A New Latin Syntax by E. C. Woodcock.