Argument in Research
Microargument is the smallest unit of proof - the individual claim and the data that support it.
- Claim: A statement of what you want to prove, what you are asserting. You must learn to
recognize claims. You must learn to write statements that clearly and succinctly state your
- Support: What you present that you believe will lead others to accept your claim as true.
You must be able to take a claim, look at it, and decide what constitutes good support for it.
Researchers must eat, drink, and sleep claims and support. Identifying a significant claim must be
like asking for candy when you were a baby. Knowing how to support a claim must come to you
like knowing what will satisfy your hunger.
The context for argument is the scholarly community
Booth, Columb, and Williams put this well. Always think of argument in terms of what you have
learned about the rules of the scholarly community conducting the inquiry - from your education;
from your reading. Graduate school is nothing more than teaching you these expectations.
What makes a good argument?
From Wayne Brockriede, "Rhetorical Criticism as Argument" Quarterly Journal of Speech 60
(April 1974): 166. The notions here are the definition of argument without regard to the context
in rhetorical criticism. Brockriede goes on to apply them to criticism in this essay, an article that
all critics should read.
"By 'argument' I mean the process whereby a person reasons his [or her] way from one idea to
the choice of another idea. This concept of argument implies five generic characteristics:
- an inferential leap from existing beliefs to the adoption of a new belief or the reinforcement of
an old one.
- a perceived rationale to justify that leap.
- a choice among two or more competing claims.
- a regulation of uncertainty in relation to the selected claim - since someone has made an
inferential leap, certainty can be neither zero not total, [always probability].
- a willingness to risk a confrontation of that claim with one's peers."
The Toulmin Model
From Stephen Toulmin, The Uses of Argument (Cambridge: Cambridge University Press, 1964),
pp. 97-100. Note that hardly anyone in Communication goes back to Toulmin to define his
model. That is partly because Toulmin himself gives differing definitions in Uses and in other
places, and partly because initial interpretations in Communication, while distorting Toulmin,
challenged old ways of thinking less. You need to read Toulmin. Disciplines will nearly always
weaken the frameworks of novel thinkers as they "adapt" them to what is familiar.
Toulmin's initial introduction of his notion of how argument proceeds is with the device of a
conversation. Think of this, in our case, as a conversation between two researchers:
- You make a claim. Toulmin would be perfectly happy with Brockriede's notion of a claim.
- Your interlocutor may accept your claim (in which case it may not have been a good claim, too
obvious). But more than likely, she will ask: "Why do you say that?" Or "What makes you say
that?" You will respond with the evidence or data that you believe will satisfy her challenge.
- Your data may satisfy your interlocutor and the conversation will end or proceed. If it does
your data warranted your claim. Notice that warranting in this conversational formulation is a
verb. It recognizes that you have chosen acceptable data in the eyes of your interlocutor. It
assumes, in other words, an argumentative field in which you are asking yourself "What kind of
data will satisfy her?" and she is asking "Is that data sufficient to prove this claim I doubt?"
Both of you resort to the argumentative field that you are a part of: a scholarly discipline, a
research tradition. Thus, in this conversational construction a warrant is not a statement (a
common misunderstanding of warrant introduced by Toulmin himself and thus not incorrect,
but confusing) but is a judgement of legitimacy, what Toulmin calls "an inference-license." The
advantage of the verb-construction of warrant is that it reminds us that a warrants are generally
implicit in data selection, and actually appear as propositions in arguments only when a
particular question is asked by an interlocutor.
(These are the elements of the Toulmin model you need to remember. But to complete his six
- Your conversation will end there if you provide warranted data to support your claim. But
maybe you misfire. Your interlocutor decides that she is your data, even if true, does not
prove your claim. She asks: "Why do you think that such data supports that claim?" Your
response is backing for the warrant.
- Or your interlocutor may say, "Yes, but you know there are assumptions you are making when
you use that data to support that claim that may or may not be true. This is rebuttal.
- Or you may respond to the challenge of your interlocutor by narrowing the breadth of your
claim and adding a qualifier to the strength of your claim.)
Presenting the Microargument
This could not be simpler. The presentation involves the following:
- Statement of the claim. This is the thesis statement of the paragraph.
- Explanation. This may be definition of terms in the claim or some clarifying illustration.
- Proof. The data that you think will satisfy an interlocutor researching in this area.
- Clinch. A restatement of the claim incorporating the evidence and/or relating the claim back to
the macroargument it is a part of.
Most often, each paragraph is a microargument. On rare occasions, paragraphs will contain
multiple microarguments. More often sufficient bulk of explanation may be required or several
proofs may be offered so that a single microargument will take more than one paragraph. In the
latter case, make certain the clinch ties the paragraphs in question together back to the claim.
Also, in the latter case, there may be an extra element in the argument - the forecast, a
metastatement that explains what you will do to prove the argument and how it will be organized.
Remember, nothing is more important than claim and support.
Macroargument is the argument of the work - an article, a book, a dissertation.
- Thesis: The claim of the macroargument. Every article, book, or dissertation will be easier to
write and prove if it has a clearly stated thesis. A clearly stated thesis is a single, simple or
complex, declarative statement that totally captures what you want to prove.
- A good thesis contains no "ands" or "buts." Thus it is not a compound sentence.
- A good thesis boldly declares your claim. Although sentences that are interrogatives or
even imperatives can serve many of the functions of a thesis, writing is easier with a
declarative sentence. The exception is some social science research in which you want a
guiding research question to serve this purpose. If so, make certain it has the other
characteristics of this thesis.
- A good thesis leaves nothing beyond it that you want to discuss. If there is material
extraneous to the thesis that you wish to discuss, you do not yet have a good thesis or
you need to reconsider including the material.
- Issue: a point of doubt. An issue results from the breaking down of a thesis in an analytic
move. Any thesis will have vulnerable points that may be the sources of doubt. You job as a
scholar is to understand the issues that are contained in your thesis.
- Stock Issues: a sort of formula that can be learned and used to reveal the issues that
may arise in a particular type of thesis. We learn for example that in a statistical
inference reliability and validity are potential issues.
- Potential Issues: an issue that could be raised in the case of your particular thesis.
Potential issues can come from the use of stock issues, or from your reading of the
literature, or from your understanding of the empirical complex of the thesis.
- Actual Issues: the issues that actually are points of dispute on a research thesis. Thus,
potential issues have become actual issues as the debate has ensued.
To Prepare a Macroargument
- Identify the issues. This is analytic work. See potential issues above.
- Assemble your proofs. Normally, of course, this involves research, usually of several kinds:
secondary literature, empirical investigation (remember empirical here does not mean
quantitative, but taking your observational and analytic powers to experience).
- Organize your argument. A well-formulated thesis usually dictates its own organization in one
of two ways. Either the actual issues become a framework for organization, or the research
method you are using to marshal your data (for example, this is a report on an experimental
study) provides you with an organizational pattern.
- Present the argument. Making decisions about what gets left out as warranted without
presentation of data, what gets placed into appendices and notes, and what goes in text,
present the microarguments to support the claim.
How does a thesis function?
A good thesis does multiple work. When you have finished your work you should be able to see
- It provides a discipline on what proof you need to provide.
- It provides Occam's Razor to decide what is irrelevant to your point and can be left out.
- It provides an organizational pattern to lay out your article.
Refuting a Research Claim
Refuting is a way of identifying why you disagree with a research claim and marshaling proof for
Types of Refutation
- Advancing the argument: Identifies an unsupported, and usually unarticulated, assumption of
the researcher you are refuting. This is called "advancing" because it creates a new issue
deeper into the research question than previously considered.
- Counterproof: Takes the researcher's claim and provides an alternative way of testing the
claim, offers alternatively-warranted data denying the claim. Thus, it provides different data,
warranting the counterclaim, that suggests the researcher's claim is incorrect.
- Identifying a fallacy of reasoning or of proof: Identifies a flaw in scientific method or the use
of fallacious reasoning that robs the data used by the researcher of its legitimacy. A frequent
strategy in heavily methodized social scientific research, rarer in humanistic research.
- Denial: simple counterdata. Accepts the warrant-license of the researcher but offers data
proving him wrong. A relatively weak form of refutation. It does not decide the issue in your
favor, merely casts doubt on the researcher's proof. Nevertheless, it can be an important form
of refutation because it suggests the probability of error.
- Questioning the argument: simply asking questions without answering them. The weakest
form of refutation. It can still diminish faith in the researcher's claims, but does not advance
Presenting the Refutation
There is a particular form which a good refutative argument takes:
- Identify the claim you will disagree with and its importance to the researcher's thesis. If
the researcher you are refuting has not stated this claim overtly, you may need to offer proof
that he, in fact, claims this.
- If you are advancing the argument, identify the assumption the researcher is making and
prove that he is making it.
- State the thesis of your argument.
- Offer the necessary explanation of your argument, particularly as it relates to the argument
you are refuting.
- Offer the proof of your argument.
- Clinch by returning to the researcher's claim and stating where that claim is now with your
position factored in.
Defending your Original Claim
A good researcher will quickly recognize the validity of sound refutation and abandon her original
claim. But, refutation may not always alter your faith in your own conclusions. These are
strategies for responding to refutation
- Deny the appropriateness of the refutation. You may argue that the assumption your critic
identified was not an assumption that you made, that you did not commit the fallacy or
methodological error charged, or that the counterproof argued does not warrant denial of your
- Advance the argument further. You may respond by pointing out an assumption made by your
critic and testing that assumption. Thus, you have raised a new issue in the line of inquiry.
- Answer the challenge. If your critic has merely pointed to things you did not provide, provide
- Offer alternative proof. Develop alternatively-warranted data supporting your claim.
- Offer additional data. The weakest form of response. It does not reestablish your claim and
you are now in a counting match with your critic.
Return to the COMM 700 Home Page