There are many ways to signal a change of direction or mood in a piece of text, but the most common is by inserting a “but”—as I’ve just done here. Alternatives such as “although,” “though,” “however,” “yet,” and “nevertheless” generally run a poor second. In fact, according to Google Books Ngram Viewer and the British National Corpus, each occurs around a tenth as often as “but.”
In research articles, though, the prevalence of “however” increases—especially in some disciplines.
I searched Elsevier’s Scopus database for article abstracts containing “however.” I then compared them with those containing “but” to bypass any differences in text length. In arts subjects, abstracts with “however” occurred around half as often as those with “but,” more precisely 54% as often, at the time of writing. In the social sciences and life sciences, it was 74% and 75%, respectively, and in the physical sciences, it was 85%. But in engineering it was 100% (1.4 million documents each for “however” and “but”).
Here’s an example from an engineering abstract, with sentences slightly shortened:
However, almost all the studies have been conducted using an actuated or electric device. However, using an electric device has several disadvantages.
The engineering enthusiasm for “however” isn’t peculiar to abstracts. I examined a sample of about 1,000 full journal articles from the physical, life, and behavioural sciences, mathematics, medicine, and engineering. In the non-engineering subjects, “however” appeared 65% as often as “but,” whereas in engineering subjects, it was 97%, within sampling error.
So, why do engineers particularly like “however”?
One possibility, suggested by a colleague, is that they’re more inclined to observe the rule that “but” should not start a sentence and they fall back on “however” instead. But as Fowler’s Dictionary of Modern English Usage points out, over several editions, the rule has no foundation in either grammar or good writing.
Moreover, in a closer look at that sample of articles, I found that engineers were little different from non-engineers in preferring “however” to “but” at the start of a sentence.
Another suggestion was that it’s not that engineers like “however” but that they dislike “but”—wherever it’s placed. They see it as less polished, less formal. Disappointingly for this idea, although “but” appeared in the sample less often in engineering subjects, so did the alternatives “although,” “though,” “yet,” and the clearly more formal “nevertheless”. It seems “but” isn’t the problem.
Rather, a simpler explanation might hold. Engineers—and others—are more likely to see “however” as a universal tool for signalling direction change, and the ready default in most circumstances. As one former engineer described it, “however” was his go-to option, something that could always be relied on.
Yet there are so many alternatives, each with a slightly different meaning and effect.
Along with those listed earlier, they include “still,” “even so,” “even then,” “even though,” “conversely,” “nonetheless,” “notwithstanding this,” “instead,” “for all that,” “all the same,” “despite this,” “in spite of this,” “that said,” “by contrast,” “in contrast,” “on the contrary,” “alternatively,” and “on the other hand” (which, incidentally, needn’t have an earlier “on the one hand” to work). Still more can be found in Quirk and colleagues’ A Comprehensive Grammar of the English Language.
Whatever your research discipline, quantitative or not, if you’re a heavy user of “however,” why not try abstinence for a while and experiment with the alternatives? Though not too many at once in case they give the reader whiplash.
If you then want to reintroduce “however,” sparingly of course, follow the advice of the journalist and teacher William Zinsser in his guide On Writing Well:
Don’t start a sentence with “however”—it hangs there like a wet dishrag. And don’t end with “however”—by that time it has lost its howeverness.
Zinsser goes on to say that you should then put it as early as you reasonably can. But as “however” draws attention to what precedes it, finding that early position can sometimes be difficult. For example, in the previous sentence, if you wanted to insert “however” (having removed the leading “But”), the natural place for it would be at or near the end, where it doesn’t work.
Luckily, so far in this piece, I’ve been able to signal direction changes solely with alternatives to “however.” I’ve avoided using it at the beginning of a sentence and at the end, and indeed anywhere between the two. As an indulgence, however, I’ve allowed it into this last sentence—which is enough.
Featured image via Flickr (public domain)
Very interesting!
I just checked a manuscript I’m working on. We have used “however” only 5 times versus 28 times the use of “but”.
But maybe I am an outlier here.😄
One problem that I find Master’s students run into is attempting to use ‘however’ as a conjunction.
This article made me laugh, not least because for the last 40 years or so I’ve been slavishly but somewhat reluctantly avoiding using “however” in mid-sentence, merely because at an impressionable age someone assured me that it was poor style. David, you’ve liberated me. Better late than never.
Was the search of the Scopus abstracts limited to sentence-initial “however”, “but” etc? Especially in mid-sentence, “but” can do jobs that “however” can’t (eg “but for the rain, we’d have gone for a picnic”). Is it possible that the content of arts abstracts makes such uses of “but” more likely? It wouldn’t explain the difference between physical sciences and engineering, though.
This nice, and yet extremely funny, little piece of guidance on technical English writing should be sent to all the Universities
in English speaking countries. When you are tired of marking huge
amount of Theses this blog calms and relaxes you very pleasantly
which is the situation I am in now.
I am a big fan of ‘But’ at the beginning of a sentence, so long as it is done sparingly. It brings emphasis. And neither do I care for ‘however’ at the beginning of a sentence. I do, however, probably do that more than I should…
Fourteen times ‘however’ in my latest paper; 8 at the beginning of a sentence. 14 ‘but’, yet only one at the beginning of a sentence. Well, well. 2 ‘nevertheless’; 9 ‘still’; 3 ‘in contrast’; 1 ‘on the other hand’ with no accompanying ‘on the one hand’. Cheers!
Oh what fun / it is to read /such a blog on do’s and don’ts, hey!
>>>>>
They see it as less polished, less formal.
>>>>>
That was my main “reason” when I wrote scientific papers.
And I also thought that capitalized “However” does look “a big change” to draw attention from readers. Now I understand that it hangs “like a wet dishrag”.
By definition, a coordinating conjunction joins two grammatically equal parts of speech–as in two nouns (milk and cookies), two verbs (ran and swam), two adjectives, or two simple sentences (phrases).
When a coordinating conjunction joins together two simple sentences (phrases), a compound sentence is the result.
When a sentence begins with a coordinating conjunction, it is not joining two grammatically equal parts of speech. Instead, it is joining an idea in the current sentence to an idea in the previous sentence. However, the joining of ideas is not the job of a coordinating conjunction.
The proper way to join an idea from one clause to an idea in the previous clause is to use a conjunctive adverb, which changes the idea in the second clause into an adverbial modifier of the verb in the previous clause.
For this reason, beginning a sentence with a coordinating conjunction (and, but, or, nor) is technically incorrect. The better alternative is to begin the sentence with a conjunctive adverb (such as “however”).
Very interesting article. Thirty-eight years ago my PhD advisor critiqued my thesis saying that I was, “very howevery” (sic). Since then I’ve tried to cut down the use of the word although I also resist starting sentences with “and’ or “but” in formal writing.
I wonder if there is a difference between native and non-native speakers in the use of ‘however’ and ‘but’. It could be that ‘however’ is preferred by non-native speakers as it has solely the function of creating the antithesis/opposition, while ‘but’ is also used for other conjuctive purposes, and therefore may seem as a less ideal/more ambivalent choice.