Much of the skill of academic writing lies in constructing a strong argument. This means establishing, supporting, and defending a position. It means being persuasive.
But the readers you have to persuade in a thesis are no ordinary readers. They are experts: scholars (and examiners) who are incredibly savvy when it comes to spotting the flaws and holes in an argument. Such sophisticated readers know what strong reasoning looks like, and they know the tricks that some writers will use to fake it.
These tricks – or ‘logical fallacies’ – are techniques used to (badly) construct an argument that doesn’t hold up to close scrutiny. Fallacious arguments can make an argument seem stronger on the surface (you’ll find a lot of ’em in internet comment sections), but they will not persuade experts or examiners. If they sneak into your thesis, they can really undermine your credibility.
There are lots of logical fallacies – but some are more common than others in a research context. So: what types of logical fallacies should you avoid when constructing an argument in your thesis?
1. Straw Man Arguments
A straw man argument attempts to make the writer’s position look more plausible by exaggerating, oversimplifying, or misrepresenting the opposing position.
Imagine that you are critiquing the work of another scholar – they have argued in favour of raising trade tariffs on certain imports, and your research supports reducing them. A straw man argument would exaggerate the other scholar’s position to make it seem absurd. Let’s say that the scholar argued for raising trade tariffs under certain conditions and within specified limits in order to boost sales of locally-produced goods. It would be false and irresponsible to ignore the scholar’s cautionary notes (conditions & limits) and rationale (supporting local manufacturers). A blatant straw man argument might disregard all this to claim that the scholar was in favour of destroying relationships with trading partners (not the scholar’s actual point).
Given that expert readers are likely to know (or even have debated) the work you are critiquing, it’s very easy for straw man arguments to be caught out. A more responsible critique of the opposing scholar would take into consideration the fullness of their position in a fair and rational way, without exaggeration. It would critique the scholar’s argument on its merits, and present an opposing view without undue disparagement.
2. Causal Reductionism
A causal reductionist argument, as the title suggests, reduces a complex phenomenon to a simple or single cause; thus ignoring the true variety of causes. Such an argument might claim that, for instance, climate change is caused by increased car ownership. That could very well be an important factor, but not the sole cause.
Researchers sometimes inadvertently adopt a causal reductionist position simply by virtue of our tight focus. If you are researching the impact of increasing car ownership on climate change, and you have ample data showing how much of an impact car ownership has, it can be very easy to forget other factors. We can become so convinced by the power of our own findings that we don’t contextualise them adequately; which sometimes leads us to lose perspective and (in the eyes of expert readers) overstate our case.
Avoiding a causal reductionist position can be as simple as ensuring that we acknowledge the importance of multiple complex causes. It’s OK to focus on one causal factor disproportionately if that’s the subject of the research, as long as that focus is appropriately contextualised. The literature review can do a lot of that work – incorporating others’ research on multiple causes for the same phenomenon can help readers to see that you, as the researcher, are aware that your work is part of a wider landscape of causal factors.
3. False Dichotomy
In a false dichotomy argument, the writer claims that there are only two alternatives when, in fact, the truth is more complex. This misleads readers into thinking they need to choose between two options, instead of allowing them to see the full breadth of options that are possible.
For example, when I try to get my preschooler dressed in the morning, I offer him two options: the blue shirt or the red shirt. This is a false dichotomy. There are many more shirts in his drawers. I am manipulating him into thinking that he only has two choices because I want to speed up the process of getting him dressed. This works on four-year-olds much better than it does on thesis examiners.
Unfortunately, some researchers fall into the trap of representing their research problem as having limited potential solutions, rather than creating a full picture of possible solutions. While it can be tempting to simplify matters for readers, it’s usually better to trust them with complexity. Is the range of potential solutions messy and complicated? If so, then describe them realistically!
4. Genetic Fallacy
An argument based on genetic fallacy ignores the merits of a particular claim and disregards it solely on its origins.
This one can be complicated — it’s not always completely unfair to dismiss a claim based on its source. After all, we are always being told to evaluate our sources in academia, and the esteem of a particular scholar, journal, or school of thought can play a big role in its perceived relevance. But that doesn’t mean that the origin of a claim is reason enough on its own to dismiss it.
This is somewhat comparable to cancel culture. For example, film scholars might prefer not to engage with the works of Harvey Weinstein or Woody Allen (and that is their choice). But it would be a genetic fallacy to deny the influence of their films on late 20th Century popular culture simply because they have since been discredited. A scholarly analysis could acknowledge their influence and their problematic aspects in tandem.
5. Post Hoc Ergo Propter Hoc
This Latin phrase translates roughly to ‘after this, therefore because of this.’ A post hoc ergo propter hoc argument is essentially a false assumption of causation based on simple timing.
For example, let’s say you’re analysing the impact of social media on political opinion. It could be tempting to track observable changes in political opinion and map them against the launch dates of particular social media platforms. But it would be too much of an assumption to conclude that shifts in political opinion in the mid-2000s were attributable to Facebook; or that shifts in the late-2000s were attributable to Twitter.
Timing alone is not enough evidence to draw conclusions. There need to be firmer links to establish a causal relationship. After all, it would be 100% true to say that I started watching more Netflix after I finished the first draft of my thesis. But did I start watching more Netflix because I finished my draft? No: that would be a post hoc ergo propter hoc assumption. The reality is that Netflix hadn’t launched in NZ before that.
If it had, my thesis would probably still be in draft.