Tracing and Refining the Inventional Topoi

The Greek word topos (Latin counterpart, locus) means literally a “place” or “region” for arguments to reside in. Although it is one of the essential tools for rhetorical invention, the evolution of its meaning and function has suffered neglect. By systematically exploring topos in a chronological order, this essay reveals its varying focal points: from formal and material categorizations in the Greco-Roman era to theological aspect in the Medieval time, from stylistic topos in the Renaissance to moral orientation in the Enlightenment, and then to argumentation scheme in the contemporary time. Through examining the related classical and modern scholarship, this inquiry finds that the two inventional tools — topos and stasis (a system of controversial issues) — richly complement each other. Focusing on the generative common topoi/topics, this study finally integrates the closely related elements of Cicero’s stasis into the popular topical system of Corbett and Connors (1999) to justify a richer and more reasonable topos model. The refined model consists of 5 major topoi/topics properly subcategorized: Definition with conviction, genus, essence, division; Comparison with similarity or difference, superiority or inferiority, more or less; Relationship with cause and effect, antecedent and consequence, opposites; Circumstance with past fact and future fact, possible and impossible; and Testimony with authority, testimonial, statistics, maxims, laws, precedents. This investigation contributes to strengthening the theoretical basis for the application of topos in the field of composition, argumentation, and rhetorical criticism.


Introduction
Topos (pl. topoi), the essential tool for rhetorical invention, has been a major concern from antiquity to the present. Scholarship on topos in the past decades has turned its critical lens to the reexamination of classical topical lore ( [14], [17], [21], [22], [26]), the comparison between classical topos and modern argumentation schemes ( [28]), and the application of topos in discourse analysis ( [12], [13]). Since topos has accrued various and sometimes overlapping definitions or forms throughout Western rhetorical history, researchers come to grips with the difficult question of its definition. Yet, it is still notoriously vague. How the meaning and function of topos evolve in different periods has received scant attention. This investigation attempts to systematically explore the dynamic evolution of topos in a chronological order and discern its fundamental patterns for justifying a feasible topos/argument model, to better serve contemporary discursive activities. Accordingly, we will focus on three research questions: What are the major features of topos in each rhetorical period? What is its general evolution trajectory? How can we establish an applicable topos model for all types of discourse? This study will add to the theoretical justification for applying common topos to various types of discourse.

The Evolution of Topos
The Greco-Roman Era: Formal and Material Topos Aristotle, who makes a significant contribution to the topical lore, is unquestionably regarded as the pioneer in the realm of topos research, although his antecedents, such as Plato and Isocrates, had been concerned with this concept. We thus trace the occidental tradition on topos from ancient Greece; the most representative works are Aristotle's Topics and Rhetoric.
In Topics (composed of eight books), Aristotle devotes his first book to the ways in which arguments are found and developed. Hundreds of dialectical topoi are illustrated within two groups: "predicable-related" and "predicable-independent" ( [26], pp. 41-42). The first category consists of topics concerned with any of the four predicables, "definition", "property", "genus", or "accident", providing "the materials with which arguments start" ( [2], p. 9). The second type of topos, deemed by Aristotle as "[the] most important to master" and "the most useful on the greatest number of occasions" ( [2], p. 115), comprises four opposites (contradictories, contraries, privation and possession, relatives), co-ordinates (e.g. "just deeds" and the "just man" are co-ordinates of "justice"), inflections ("justly" for "just"), the greater and lesser, and the like degree. Aristotle designates this kind of topics as the principles of constructing or demolishing theses containing any of the four predicables; accordingly, the knowledge of them "secures a basic set for approaching the construction of virtually any argument" ( [26], p. 42). Aristotle's another work concerned with topos is his Rhetoric, in which he distinguishes rhetorical topoi between common (koinoi) topoi and special (idioi) topoi. These two types are regarded respectively as "formal" and "material" categories by modern scholars ( [17]), who embrace the view that special topos offers the material for propositional statements and the common type offers forms of inferences into which the material may be put. Aristotle seems to give more attention to common topos, as twenty-eight formal topoi are elaborated, from which rhetorical arguments can be drawn: proof based upon opposite ("temperance" and "licentiousness"), key-word modification ("just" does not always mean "beneficial", or "justly" would always mean "beneficially"), correlative terms ("where it is right to command obedience, it must have been right to obey the command"), "a fortiori" (more or less), definition, division, cause to effect and so on ( [1], pp. 142-155). These twenty-eight rhetorical topoi share cozy link to rhetorical syllogism or enthymeme, as Aristotle specifies topos by "an element of enthymeme", the same thing as "a line of enthymematic argument" ([1], p. 163). For example, topos of opposite clearly functions as major premise in the following syllogism, "If licentiousness is hurtful, and temperance is the opposite of licentiousness, beneficial is the opposite of hurtful, then temperance is beneficial." The above "predicable-independent" group, definitely belonging to common topos, also functions as a crucial element of rhetorical syllogism. In sum, both Aristotle's dialectical and rhetorical topos not only bring the process of inference into focus, but also provide principle for warranting the movement from premise to conclusion ( [22], p. 26).
In Roman rhetoric the distinction between common topos and special topos is left undeveloped, so is its inferential function. Two inventional systems are advocated by Roman rhetoricians to find materials for constructing arguments: loci/topoi and stasis. The topical system commences with Cicero's discussion of subject-matter in De Inventione I ( [9], pp. 31-43), from which he identifies a list of common topics, the attributes either of the "person" or of the "act", and treats them as support for all the arguments (p. 34). It is in his mature works of De Oratore and Topica that Cicero delineates a revised system of loci, which deletes the topics concerned with attributes but places more emphasis on causal reasoning like the form of "if…, then…". Nineteen loci are subdivided into two groups: intrinsic arguments attached to the subject under discussion, and extrinsic arguments "depend[ing] principally on authority" ( [9], p. 397). Arguments related to the subject adhere very closely to Aristotle's common topos, such as arguments from definition, division, etymology, and circumstance (genus, species, similarity, difference, opposite, antecedents, consequents, causes, effects, and comparison). In an apparent departure from Aristotle, loci elucidated by Cicero neither involve the four predicables, nor make reference to the inferential strategies connecting propositions of rhetorical syllogism or enthymeme. Loci communes, or "common places" in Ciceronian sense, which are material and "ready-made arguments" ( [26], p. 107), take on the character of static lines of arguments rather than a dynamic system of topic discovery in Aristotle. The focus of topical system, as mentioned by Leff, "shifts from the discovery of inferential connectives to the discovery of materials for argument" ( [22], p. 29).
Following Aristotle and Cicero, Quintilian in The Orator's Education calls topoi/loci "Places of Arguments" and defines them as "the areas where Arguments lurk and from which they have to be drawn out" ( [25], pp. 375-377). Positioned under deliberative, forensic and epideictic discourse, topos discussed by Quintilian is related to the four predicables, similar to those of Aristotle: "genus", "species", "property", and "difference", which indicates its relation to the form-oriented common topos. On top of this similarity to Aristotle's dialectical topical system, topics bearing resemblance to Aristotle's rhetorical topics and Cicero's nineteen loci are explicated in each discourse, such as arguments from definition, conjugate, comparison, relationship, and authority.
The first stage of topos from Aristotle to Quintilian reveals its various type and different functions. While Aristotle attaches much importance to inferential strategies implied by topos, in Roman period, topos more distinctively functions as ready-made search list that can repeatedly be used to similar cases.

The Medieval Time: Theological Topos
The topical systems in Roman times undergo further transformation in the Middle Ages. In the early medieval period, Augustine in De Doctrina Christiana (composed of four books) treats topoi as "hermeneutical" guides for explaining the Scriptures. Books One to Three concern the comprehension of God's Word. The rules such as syllogisms, definition, division, and classification greatly help people understand the theological truth ( [3], pp. 117-119). Focusing on the tropes that function in interpreting ambiguous signs, Book Three specifically dwells on the figures of speech that show a close association with topoi in the Scripture. Figures of irony and antiphrasis are similar to the topic of opposite (p. 173); the trope of synecdoche encompasses the argument of part for whole or vice versa (p. 187). Since Christians traditionally resort to the Scripture for inventional materials, Book Four concerns how to use the understanding of theological truth to persuade the believer. Therefore, topos, oriented towards law court or ceremonies in classical invention, derives from the interpretation of sacred materials in the Medieval time.
When it comes to the inventional system in the late medieval period, topos expounded in Boethius's [6] treatise De Topicis Differentiis sees a remarkable resurgence of inferential principle advanced by Aristotle. A Topic, as Boethius considers, "is the seat of an argument" (p. 30). Instead of relying on predicables for topical categorization, Boethius comes up with two forms of topics. One type is "maximal proposition" -"the foundation of an argument" (p. 47)which plays a role in providing inferential connectives for topical relationships such as whole/parts and cause/effect; the other is topical differentia, which groups these maxims. Twenty-eight differentiae are classified into three divisions according to the relationship with subject at issue: intrinsic, intermediate and extrinsic, among which some subdivisions are the same as Aristotelian common topoi, such as definition, whole or genus, causes and consequents, similars, the greater and the lesser, and opposites. In the view of Boethius, topics are essential tools aiding in "the investigation of truth" (p. 42).
Another typical feature of rhetorical treatises witnessed in the late medieval period is the increasing emphasis on the arts of preaching, letter and poetry writing, indicating that the center of rhetoric is drifting from matters of argument into those of arrangement and style.

The Renaissance: Stylistic Topos
The scope of classical rhetoric was pruned in the Renaissance. Since Peter Ramus has removed invention from the five canons (invention, arrangement, style, memory, and delivery) to dialectic in his Institutes of Dialectic (1543), there appears to be a shift of emphasis from invention to style and logic.
In Shakespeare's Use of the Arts of Language, Sister Miriam Joseph [19] dwells on how Shakespeare's figures of speech are closely linked to topos. Part Two of her book offers a list of logical topics to which the figures are related, such as definition, contrary, comparison, and cause and effect. A large section is allotted to explain the relations between tropes and topos in Part Three. Figures of litotes, paradox and irony, she claims, show an intimate connection with topics of contrary and contradictory (pp. 323-325); metaphor, "a trope based on similitude", shares a cozy relation with topic similarity (p. 328).
In 1553, Thomas Wilson publishes his representative work The Arte of Rhetorique, in which he reiterates the topics discussed by Aristotle and Cicero. In order to authorize arguments, an orator, to a large extent, must resort to the places of logique, defined by Wilson [31] as "good occasion to finde out plentifull matter" (p. 36). In addition, this occasion helps set forward the process of "prouing the matter, and searching out the substance or nature of the cause" (p. 117). Some of the six places of logique, definition, causes, parts, effects, things adioyning, and contraries, are topics in a broadly Aristotelian and Ciceronian sense.
The shift of focal point from rhetorical invention to style and logic results in the decreasing of monographs concerned with topos, compared with the Greco-Roman era and the Medieval time. In the Renaissance, rhetoricians implicitly explain the arts of invention via combining figure of speech and inventional topos.

The Enlightenment: Moral-Oriented Topos
In the wake of the Renaissance, rhetorical works in the Enlightenment appear a tendency to adopt knowledge acquired from human experience as persuasive strategy. Conforming to the intellectual environment characterized by rational science in which he lives, George Campbell [11] in The Philosophy of Rhetoric uses evidence -"the foundation of all conviction and persuasion" (p. 61)rather than topos as the core concept of logical persuasion. He relates reasoning to different faculties of understanding, imagination, emotion, and will. It is thus undoubted that Campbell neither advances the reasoning of enthymeme nor the classical inventional systems, as both scarcely engage human nature (xxv). Instead, statements or the starting points of arguments must be authorized, Campbell asserts, "by sensation and by a logic of belief that has sense data" (xxxi). In this sense, different sources of evidence he describes are subdivided into intuitive and deductive; the former kind consists of three types: "pure intellection", "consciousness", and "common sense" (pp. 35-42), with which two processes of reasoning under deductive evidence start, demonstrative and moral. Campbell places higher value on moral reasoning, for this kind of reasoning supported by human experience is more useful in acquiring new knowledge through inference from particulars to general (p. 43). The "tribes" under moral type encompass experience, analogy, testimony, and calculations of chances (pp. 50-58), belonging to topoi, special or common.
How things can be proved also draws the attention of Richard Whately, another leading Enlightenment rhetorician. In Whately's view, rhetoric is "argumentative composition", argument being the basis of persuasion ( [30], pp. 4-6). In his representative work, Elements of Rhetoric, Whately lays down two great classes of arguments: one is argument from cause to effect; the other includes all the remaining arguments (p. 46). The detailed discussion regarding arguments from testimony, progressive approach, example, induction, experience, and analogy accounts for a large section of this book. Apart from the argument sources sharing great similarity to those of Campbell, evidence for demonstrative and moral reasoning continues to be a major point taken seriously by Whately. However, he contradicts Campbell's assertion that logic "has not the least affinity to moral reasoning" ( [11], p. 62). From Whately's point of view, these two kinds of reasoning bear similar relationship with Aristotle's common topos and special topos, as moral reasoning provides materials for statements, and demonstrative type presents inferential form connecting those materials.

The Modern and Contemporary: Argumentation Scheme
From the late nineteenth century up to now, topos seems to have disappeared. But a careful look at the works of this period reveals that inventional topos does not fade away, for it either evolves into a different form or is implicitly incorporated into other theories, like Kenneth Burke's dramatic pentad of act, scene, agent, agency, and purpose. Not until the mid-twentieth century have the argumentation schemes emerged that reiterates the inventional function of topos.
The two most influential treatises of argumentation in modern times are Stephen Toulmin's The Uses of Argument and Chaim Perelman's The New Rhetoric: A Treatise on Argumentation (coauthored with Olbrechts-Tyteca). Toulmin [27] in this monograph puts forward a model of argumentation, which comprises three basic elementsdata, warrant, and claim (complemented by backing, rebuttal, and qualifier)functioning as minor premise, major premise and conclusion respectively. In this argumentation model, warrant takes the place of classical topos as major premise in syllogistic structure, for both of them "link ... the data from which an argument begins and the conclusion with which it ends" ( [21], p. 220). Due to the practicality and operability, this model has been widely adopted by modern scholars as an analytical tool in argumentation. Derek Ross ([26], p. 9), for example, in the light of the complicated argumentative steps of Toulmin Model, shapes a simplified scheme of communication, in which the "commonplace" is tactfully used as "discursive shorthand" functioning in moving the audience "to a place of shared understanding" at any point in the argument process, thus minimizing the steps of argumentation to a large degree.
In the sense of Toulmin, his six-part argument scheme is "field-independent", through which people can argue in different contexts (or fields). With more attention paid to real context and natural language, Toulmin's inference from data to claim seems outside the scope of formal logic. Yet, it is a rather elaborate form of Aristotle's syllogistic structure, which still focuses mostly on the validity of reasoning without any attempts to take audience or their values into consideration. Belgian philosopher Chaim Perelman thus advances "New Rhetoric", a continuation of Aristotle's rhetorical argument, of which the essence is a theory of argumentation. But unlike classical argument theory, argumentation in new rhetoric is characterized by winning "the adherence of the members of an audience", Perelman [24] argues (in The Realm of Rhetoric, a concise version of The New Rhetoric), rather than aiming to "deduce consequences from given premises" (p. 9).
In The New Rhetoric, Perelman's topical system lies in two major categories: one is called the starting points or premises of argumentation accepted by the audience, covering such sub-classes as facts, truths, presumptions, values, value hierarchies, and loci; the other category is embodied in the forms of argument or techniques of argumentation (structures moving from starting points to conclusion). Regarding the first type, loci mentioned in new rhetoric refer to "premises of a general nature that can serve as the bases for values and hierarchies" ( [23], p. 84) rather than "storehouses for arguments" in classical rhetoric. Shifting to the second manifestation of topos, loci/topics are delineated from two processes of argumentation: loci from association (bringing separate elements together) and disassociation (separating previously tied elements); the former includes three sub-types: quasi-logical argument, "similar to the formal reasoning of logic" ( [23], p. 193), arguments based on the structure of reality (sequential relations and the relations of coexistence), and the relations establishing the structures of reality (establishment through example, illustration, analogy, or metaphor). Several argument patterns of the associating technique identified by Perelman and his co-author resemble Aristotelian common topics, to a large degree, such as arguments from definition, contradiction, comparison, and causal link ( [28]).
The study of argumentation schemes goes one step further with more attempts to find commonly used forms of argument. Fahnestock and Secor [15] in A Rhetoric of Argument illustrate tactics for argument invention; for instance, the way to claim the nature of things consists of definition, comparison, disjunction, and causal arguments.
In Public Speaking as a Liberal Art, Wilson and Arnold [29] set out a 16-point scheme of review topics applicable to all subjects, such as existence, degree, time, causality, and correlation. More recently, specific types of inference and its corresponding warrants such as parts to whole, comparison, correlations, cause and effect, and testimony are elucidated by Zarefsky [33] in the textbook Argumentation: The Study of Effective Reasoning.
Confronted with an increasing demand for arguments in diverse fields, modern rhetoricians reach a consensus that it is requisite to reorganize universal topos so as to establish a more scientific and operable topical system for contemporary argumentation activities. Therefore, another typical evolution of topos in the contemporary era is the hierarchical synthesis of topics. Below outlined are two most important and relevant approaches of topos categorization.
Richard Weaver in his representative work, "Language Is Sermonic" (1953), selects some topics from Aristotle's Topics to construct an applicable topical system. The sources of argument cover four categories which are ranked according to their ethical worth: definition, similitude, cause and effect, testimony and authority, among which the last type is from the external source. Of the internal three, the first kind which deals with the nature and being is the most ethical and persuasive type in Weaver's view, while cause and effect is the lowest in persuasion, especially its subtopos of circumstance which is considered "the least philosophical of all the topics of argument" and employed only "when we don't know anything else to plead." ( [5], p. 1356) Another representative list of topoi is summarized by American rhetoricians Edward Corbett and Robert Connors [12] in their very renowned Classical Rhetoric for the Modern Student, where they refine Aristotelian common topoi into five general types bearing much similarity to Weaver's system both in classification and sequence or hierarchy: definition, comparison, relationship, circumstance, and testimony. Testimony is continually regarded as external source in the view of Corbett and Connors. Each type contains rather clear subdivisions: under "definition" are genus and division; similarity, difference and degree are included in "comparison"; "relationship" consists of cause and effect, antecedent and consequence, contraries, and contradictions; possible and impossible, past fact and future fact belong to "circumstance"; and the external source "testimony" comprises authority, testimonial, statistics, maxims, laws, and precedents (examples).
This brief sketch of topos from ancient Greece to the contemporary age suggests not only its different meanings accrued but also diverse purposes accumulated. However, whether as "place" storing arguments, "hermeneutical" guide, art embellishing language, evidence embodying human nature, or analytical tool for argumentation, topos invariably functions as inventional device and remains a focal interest of rhetorical researchers.
So crucial are the topics in rhetoric that it is not difficult to detect a tendency to synthesize feasible topical systems for discourse construction and analysis. The aforementioned two models, although representative and popular, are not comprehensive and applicable to some extent. For one thing, Weaver fails to offer clear or reasonable sub-types for each topos; treating circumstance as a subvariety of cause and effect seems not justifiable. For another, the subdivisions of each topos in Corbett and Connors, though illustrated with typical examples, are not justified theoretically, and some of them are not adequately classified, such as the sub-types of "definition". Be that as it may, taking into consideration the evolution trajectory of topos, we prefer to refine the topos model by Corbett and Connors, for the following three reasons.
First and foremost, this layered system not only inherits those favored common topics passed down from Aristotle, but also reflects the conciseness and prominence highlighted in modern and contemporary times. Second, these two acclaimed rhetoricians have already adopted the five general topoi as powerful explanatory tool in argumentation analysis, evidently manifesting the operability and feasibility of their model. In the meantime, Classical Rhetoric for the Modern Student where their topos model is included has been extremely influential (as evidenced with four editions), which, to a certain extent, proves the validity of their model.

A New Topos Model Refined with Stasis
Given that the notion of stasis shares a close affinity with topos as both serve vital roles in invention, this investigation tries to justify a more reasonable and substantial topos model with the related elements of stasis.

Relation between Stasis and Topos
Stasis (status or constitutio in Latin), rivaling topos in rhetorical invention, is treated by Roman rhetoricians as means for arguments discovery. The term has been assigned a variety of meanings. In De Invenione I, Cicero [9] calls stasis "question" or "issue" and defines it as "the first conflict of pleas which arises from the defence or answer to our accusation" (p. 21); later scholars strive to offer more succinct definitions, for instance, Fahnestock and Secor [16] regard stasis as "a taxonomy, a system of classifying the kinds of questions that can be at issue in a controversy" (p. 137). And this "systematic way of asking questions about rhetorical situations" is efficient for rhetors to "determine just where it is that the disagreement between themselves and their audience begins" ( [13], p. 53), thus "[giving] direction to the entire investigative process" ( [20], p. 27). Therefore, stasis serves a vital role in invention; it also plays a significant part in "decid[ing] on a thesis" ( [12], p. 28). On top of the core function of rhetorical invention, the system of stasis has been employed effectively in rhetorical criticism ( [13], [32]).
Since both stasis and topos are commonly recognized as two significant means of invention, these two doctrines are richly complementary to each other. The sinewy connections between them can be traced back to antiquity if we give a closer look at the classical monographs. In Topica, for instance, Cicero [9] defines "a topic as the region of an argument, and an argument as a course of reasoning which firmly establishes a matter about which there is some doubt" (p. 387), where "doubt" implicates issue or stasis ( [32], p. 104). And Rhetorica ad Herennium, the earliest rhetorical manual, establishes an inventional system integrating stasis and topos. When discussing deliberative speeches, the author primarily divides the issue, for example, "security" or "honor", into two kinds: "either of the kind in which the question concerns a choice between two courses of action, or of the kind in which a choice among several is considered" ([10], p. 157). This classification is obviously stasis-oriented, determining the direction of invention, upon which appropriate topics are then selected. Connections between topos and stasis repeatedly occur in modern rhetoric as well. Brockriede and Ehninger [7] link artistic proofs with four types of "disputable questions" (staseis) and claims in Toulmin Model. For instance, definitive claims which relate to the question of definition are supported by arguments from parallel case, analogy, or authority (p. 52).
More distinct explanations of the firm relations are put forward by George Kennedy and Janice Lauer. Kennedy [18] claims that Hermagoras's "theory of stasis ... restated Aristotle's theory of topics as a system of categories adapted to the law courts, which a student of rhetoric could memorize and apply to any situation" (p. 88). In Lauer's [20] well-accepted Invention in Rhetoric and Composition, the link is described in the same line: "... status as initiators of discourse; special and common topics as exploratory arts" (p. 22).

Cicero's Stasis System
Stasis, like topos, encompasses different types and subdivisions. The origins of this classical theory are attributed to Hermagoras whose work fails to survive, except for the discussion in Cicero's De Inventione, De Oratore, Topica, Quintilian's Institutio Oratoria, and other treatises concerning rhetorical invention. Hermagoras divides stasis into four types: conjecturalis, when the "fact" is at issue; definitiva, whether an admitted action falls under the legal "definition" of a crime; generalis, the issue of the "quality" of an action, and translatio, objection to the legal process or "transference" of jurisdiction to a different tribunal. According to Cicero [9], "every subject which contains in itself a controversy to be resolved by speech and debate involves a question about a fact, or about a definition, or about the nature of an act, or about legal processes" (p. 21). From Hermagoras to Cicero, the canonical four staseis concisely put are "fact", "definition", "nature", and "legal process"/"action" in Cicero's later work. Besides this layer, there are rich subdivisions under each type which are elaborated in greater detail in Cicero's De Inventione, De Oratore and Topica.
For the first presented "fact" stasis, substaseis assume a wide divergence among Cicero's treatises. In De Inventione, this type of stasis concerns questions about "what has been done? ... and what is being done? ... and what is going to occur" ( [9], p. 23). Cicero evidently subdivides "fact" into triple categorizations as "past", "present" and "future". It is in his mature works of Topica ( [9], p. 445) and De Oratore ( [8], p. 91) that he offers more acceptable divisions which can be tersely summarized as "existence" (Does it really exist?), "origin" (What is its origin?), "cause" (What causes it?), and "change" (Can it be changed?).
As for the subelements of "definition", De Oratore and Topica still give more detailed explanations (than De Inventione). In the former work, Crassus (Cicero's spokesman) states that "disputes as to definition arise either on the question of what is the conviction generally prevalent, ... or on the question of the essential property of something, ... or when a thing is divided into parts, ... or on the problem of defining the special form and natural mark of a particular thing, ..." ( [8], p. 91). Four categorizations are distinctly embodied in these lines: "conviction", "essence", "parts", and "mark". All these elements possess wide application value in discourse construction and analysis. In Topica ( [9], p. 447), Cicero uses the identical subdivisions of "definition".
With the stasis of "nature", Cicero's later works show particularly applicable sub-types. In Topica, the question about "nature" is "put either simply or by comparison; simply as in the question: Should one seek glory?by comparison, as: Is glory to be preferred to riches?" ( [9], p. 447). Cicero further illustrates how to argue simply or comparatively: "sought" or "avoided", "right" or "wrong", "honorable" or "base" for the simple way/direct judgment; "sameness" or "difference", "superiority" or "inferiority" for the comparative judgment ( [9], p. 447). The same ideas regarding "nature" also occur in De Oratore ([8], pp. 91-93).
The last stasis of "action" presents even more divergences, in De Inventione, subdivisions can be summarized through Cicero's discussion ( [9], p. 33) as "person", "manner", "court", "law", and "time". However, in De Oratore, things are quite different. It is elucidated that "those referring to conduct [action] either deal with the discussion of duty ... or are employed either in producing or in allaying or removing some emotion" ( [8], p. 93). Two concise subelements can be extracted from these lines: "duty" and "emotion management". In Topica ( [9], p. 449), Cicero employs similar subelements of "action" in De Oratore.
The inquiry of substaseis continually grasps the attention of contemporary researchers, for the purpose of discourse construction and analysis. On the basis of Cicero's categorizations, Crowley and Hawhee [13] raise a list of questions under each major type of stasis to apply them to analyzing texts. More constructively, Yuan, Harris and Jiang [32] perform a systematic exploration of Cicero's system of staseis; while maintaining most of Cicero's mature (sub)categories, they establish a modified stasis system by integrating Kenneth Burke's dramatic pentad of act, scene, agent, agency, and purpose to replace Cicero's elusive "action" subdivisions, as shown in the table 1.

A Refined Topos Model
The interpretation of Cicero's stasis system and its sub-types provides more evidence to affirm the tight connection between topos and stasis. Upon this connection and the survey of topos evolution, we are to justify a richer and more reasonable topos system. As mentioned before, in the history of topos studies, Corbett and Connors' system, comparatively speaking, looks most clear and applicable, but in the light of stasis system, it can be further refined into a clearer and more operable model. Now before refining, let us first reexamine their system in the Figure 1 designed upon their listing of "The Common Topics" and the corresponding illustrations ( [12], pp. 87-121).
Generally, we tend to accept this three-layered topos system. The first layer with the division of "Internal" and "External" is also affirmed by Richard Weaver as introduced previously, and it can be traced back to Aristotle's artistic and non-artistic means of persuasion in his Rhetoric (Book I, Chapter Two). Also acceptable is the second layer of five major common topoi, since they are the prominent ones frequently mentioned in topos history and are at the same logical level. With regard to the third layer, we think that the subtopics under the external "Testimony" are clearly and comprehensively provided, no need for adjusting; however, the subdivisions of the four internal topoi are not adequate or reasonable enough, so we will exert our efforts here for the focused refining.
Inspired by the related elements in Cicero's stasis system, for Definition, the subtopics better start with Conviction (sound description of the concept), then Genus (its belonging class), and then Essence (its essential property) is expected before Division (parts in Cicero's system). Although there is also Mark under definition stasis, not all concepts have distinctive outer marks, so we can omit this one and justifiably add two elements of Conviction and Essence to Corbett and Connors' Definition subtopics. For Comparison, closely related to Nature stasis, we find that their first two subdivisions (Similarity, Difference) are actually of one dimension, according to comparative substasis (see Table 1) which also includes another dimension of Superiority or Inferiority mentioned too by Corbett and Connors ([12], p. 92) but unduly neglected in their system. Degree sub-type should be kept, as it is reasonably inherited from Aristotle who interprets it as "more or less" and "greater or lesser". And it can be interpreted in various ways, such as "longer or shorter" and "stronger or weaker", but in correspondence with the other two sub-types, we can use the first pair to stand for all the other similar cases. In sum, the topos of Comparison can be reasonably adjusted to Similarity or Difference, Superiority or Inferiority, More or Less.
For Relationship, their subdivisions are basically acceptable. The first one, Cause and Effect is definitely necessary, as it indicates the causal relation. The second, Antecedent and Consequence, indicating the temporal or conditional relation, is a necessary dimension as well. The last two (Contraries and Contradictions) actually belong to the same oppositional relation which, according to Aristotle's Categories, include four particular kinds: contradictions (e.g. "he is sittinghe is not sitting"), contraries ("the good and the bad"), privation and possession ("blindness and sight"), relatives ("the double and the half") ( [4], p. 19). So, for Relationship, we keep the first two of Corbett and Connors' subdivision and replace their last two with the more inclusive Opposites.
For Circumstance, Corbett and Connors include two subdivisions: Possible and Impossible; Past Fact and Future Fact. According to their interpretation, the first one revolves around action possible/encouraged or not and the second one concerns whether or not something has happened or will happen. The two subdivisions actually relate respectively to Action stasis and Fact stasis. Then, we see the logical problem in their arrangement of the sequence: definitely fact has to be investigated before considering action. So, we need to reverse their order. With regard to the in-depth survey of fact or the proper judgment of action possibility, we can rely on the substaseis of Fact and Action (see Table 1). For conciseness, we will not list them out in the following figure. Figure 2, with the same cast as Figure 1, presents our refined model of the common topoi/topics in a more concise and direct way, which is easier to locate the adjusted parts.
It should also be pointed out that the sequence of the first and second layers from left to right does indicate certain hierarchy of persuasiveness, according to master rhetoricians, such as Aristotle and Weaver, but this sequence is not necessarily the order of topics' appearance in the actual use, and the five common topics may not appear together in one passage, not to say all the subtopics. This revised model provides a justifiable framework of the possible active common topics to be chosen from in varied discursive activities, and rectifies, to a certain extent, the overemphasis on some external topics, and the unduly neglect of significant internal topics, especially that of Definition.
Due to the limit of space, this refined new model will not be applied to discourse analysis, but it is definitely solid and applicable. For Corbett and Connors have succeeded in applying their topos system to the analyses of various genres, such as Socrates' "Apology", Edmund Burke's "Letter to a Noble Lord", Thomas Huxley's debate regarding "Science and Culture", and Rachel Carson's deliberative discourse, "The Obligation to Endure". Evidently, our refined model, a justified modification of theirs, will prove no difficulty in applicability.

Conclusions
The dynamic evolution in the first half of this article clearly reveals that the definitions, referents and roles of topos are richer and more complicated than it is usually represented. The general evolution trajectory sketches out two forms: classical topos and modern argumentation scheme. With its role varying in such different ways, traditional topos is more than a "place" where arguments reside: it can function as "hermeneutical" guide, find middle terms connecting syllogistic propositions, be incorporated into figures of speech embellishing art of language, take the form of evidence reflecting human nature, or serve as explanatory tool in discourse analysis. There has been tight relation between these two general topos forms, as modern argumentation schemes can trace its origin from classical topoi, but puts more emphasis on reasoning in natural contexts.
Via exploring topos in each rhetorical period, we find that some types of topos -Definition, Comparison, Relationship, Circumstance and Testimonyrecur and remain vibrant throughout the evolutionary process, but they can be further enriched or clarified, especially for the first four. Through examining related classical and modern inquiries, the sinewy connections between two inventional toolstopos and stasishave been detected. This paper, thus mainly integrates some subelements of Cicero's well-accepted stasis system to refine the popular topos system of Corbett and Connors [12], so as to construct a richer and more reasonable topos model that can be applied to composition, rhetorical criticism, argumentation studies, etc.
Confining our research in common topos, applicable to all subjects, does not suggest that the special type, unique to specific disciplines, bears no value. Instead, recent scholarship on special topos witnesses a spiral increase, as this context-driven approach to argumentation is a more direct as well as efficient way in shaping attitudes or values. New trends towards topos research have emerged from time to time. In the proceedings Topic-Driven Environmental Rhetoric [26], topics have been tactfully used to frame environmental discourse and form corresponding ideologies, via both common topos/topic and special topos/topic, or even the integration of them. This new topic-driven approach, to a certain degree, will bring a promising future for topos research in different fields.