Commonplaces should, if possible, be avoided. Don’t preface your introduction with sentences like “English grammar has been a fascinating challenge to linguists for years” or “Speaking a language is what makes us human”. That may or may not be true, but it does not further your project at all, and such pre-beginnings become increasingly annoying the more of them you have to read.

Academic texts are not primarily meant to “sound good” — not at the cost of the contents. Don’t use words if they don’t mean what you want to say. For instance, many students overuse the word moreover, probably because they were told to connect their sentences somehow. But moreover means that what follows adds more support to an argumentation already in progress – don’t use moreover if you are going to introduce a completely new point. Another expression overused these days is delves into. If I got a Euro each time I read this in a term paper, I’d be able to invite you all out for a very nice dinner. If you do want to use it, make sure to use it correctly. Hypotheses don’t delve into things, and neither do analyses. People or texts delve into things.

Consider the logical connections in the following passage and evaluate the use of conjunctions and adverbials. How could the passage be improved?

Studies show that the frequency of use of phrasal verbs has decreased in academic genres (Kovács 2008; Leone 2021; Rodríguez-Puente 2019; Thim 2012). In addition, Thim (2012) and Rodríguez-Puente (2019) refute the assumption that phrasal verbs are informal and colloquial. Thim (2012) points out the lack of evidence for the alleged informality of phrasal verbs while Rodríguez-Puente refers to Thim's (2012) claim "that their alleged colloquial character is a mere 'conspiracy'" (Rodríguez-Puente 2019: 283), but later concludes that "phrasal verbs have a more colloquial character than their one-word Latin equivalents" (Rodríguez-Puente 2019: 283). Contrarily, Kovács' (2008) research demonstrates that phrasal verbs are neutral rather than informal.

Leone (2021) further investigates the shift from using phrasal verbs to using simple verbs in court rooms. Moreover, the study shows that certain phrasal and simple verbs can be used interchangeably in the same context while others cannot. Some single-word verbs and multi-word verbs can be used interchangeably if they do not differ in meaning and usage while others cannot (Kovács 2008:9).

The conjunction but in the middle of the passage, and possibly contrarily two lines later, are the only one that are used well. The first sentence is about the frequency of phrasal verbs and the second sentence is about style, so connecting them with moreover is confusing. If Rodríguez-Puente (2019:283) agrees with Thim’s claim that phrasal verbs are not necessarily colloquial (it would have been better to make that clear instead of just saying that she “refers to” the claim), the sentence should not begin with while, which expresses contrast. (And if Kovács demonstrates that that phrasal verbs are neutral rather than colloquial in 2008 already, the point ought to be settled. Why do Thim and Rodríguez-Puente argue about this again later?) Saying that Leone “further” investigates the shift from using phrasal verbs to using simple verbs suggests that they investigated it before and are not following up on their previous investigation. This is probably not what is intended. Since a change in usage patterns/relative frequencies is a new point not directly connected to the style question (hence the new paragraph), there is no need to use a connector at all. The following sentence gives the results of Leone’s study; beginning it with moreover is, again, confusing, since it makes us expect a new point in support of a previously supported claim. (It does seem as if Leone’s study makes at least two very different points, but if the first sentence tells us what the study “investigates”, we assume that this is its general topic, and expect the following text to inform us about the findings.) These repeated inaccuracies make the passage difficult to read.

The problem, probably, was that the author was told to read a particular number of texts on the topic and collected bits and pieces at random before having an agenda of her own, and then tried to work unconnected pieces of information into a connected text, which resulted in cohesion without coherence. Instead of trying to incorporate as many quotations as possible, develop your own argument first and then add support or criticise other views where it makes sense in your own line of reasoning.

It is good to let your own style of writing be informed by the style of the academic publications you read. When we use language, we always work with material we hear or read, manipulate it and send it on. But don’t sacrifice sense to fluency. It’s not nice to say that a study “aims at” doing something. That suggests that it doesn’t succeed. Don’t say that just because you feel that aims at sounds good. The same goes for focuses on – don’t say that a study “focuses on” something if that something is that study’s only object of research. Just say investigates or explores. If you feel tempted to write that something “shines a light on” something, “highlights” something or “underscores” something, try to be precise instead. Does it explain something, or demonstrate something, or merely draw attention to something? And don’t overuse words like subtle and nuances. Repeated use of such terms makes your text superficial and may suggest to your reader that you were unable or unwilling to work through the details.

In general, you should always write in a goal-oriented way. You need to argue your point in a very limited number of words, so say what you have to say concisely, and use simple words if you can. The complexity of the text should not exceed the complexity of the contents. Einstein said (supposedly): “If you can’t explain it simply, you don’t understand it well enough.”

The author is dead. 🙂

In literary studies, we say that the biography of the author is irrelevant to the interpretation of a text because the author either managed to say what they meant to say, in which case we should be able to find it in the text itself even if we didn’t know anything about the author, or they didn’t manage to say it, in which case it doesn’t matter. (Umberto Eco said this, I think.)

It’s the same with language. If there’s something in history but the data don’t show it, it’s irrelevant. If there’s something in history and the data show it, too, you don’t need the history.

Granted, in a paper on Latin loanwords in Old English, it does make sense to mention that Britain ceased to be a Roman colony about 100 years before the coming of the Anglo-Saxons, and that the Anglo-Saxon kingdoms in Britain were christianized starting in the sixth century, Latin being the language of the church. But you don’t need an extra section for it, mention it in the introduction if you like. More importantly, see if you can confirm the story with your data: do the Latin words in Old English transport Roman concepts, or are they Christian terms? You can speculate about what the contact situation is likely to have been after your analysis, on the basis of your findings.

Or take language contact with French. (The topic of Norman-French loanwords in English is banned, by the way.) Of course it’s never a bad idea to remind people of the Norman Conquest in 1066 (one sentence!), but rather than begin with “French was the language of the upper class” (How do you know? How does your history book know?), look at the loan words themselves and see what they tell you about who was or wasn’t upper class. This doesn’t belong into a “Historical Background” section, it’s in the foreground, it’s your main research interest, in fact.

Final example: Scandinavian place names and switching code. The whole point of investigating non-cognate sound substitution is to find out whether Old Norse and Old English ought to be considered different languages or not. You cannot present a tree diagram and say “look here, Old Norse is in a different branch from Old English, so they are different languages”. People draw these tree diagrams based on studies of language. They illustrate our best guess, they are not independent information. Your job is to present linguistic evidence and confirm or revise the diagrams.

Term papers are argumentative, not narrative texts. Purely narrative passages will not win you any points. You are writing about the history of English, not of England. So, whenever you’re tempted to preface your paper with a “Historical Background” section, make sure it contains only the information really necessary to understand the rest of the paper, and see if it really needs to be a separate section. You don’t have space to waste.

I’ve repeatedly found myself thinking “this could turn out to be an excellent term paper”, and then, when turning the page, being shocked to see that the next headline was “Conclusion”. Where did the analysis-bit go?

Ideally, the main part of your paper (i.e. the analysis of actual language, including discussion/interpretation), should make up about two thirds of the total words of your paper. So if you’ve finished your introduction and you’ve written 1000 words already, something’s wrong.

The good news is: coming to the point fast is a skill that can be learned. See for yourselves, which passage would you prefer to read?

In order to find out whether phrasal verbs occur in academic English, I queried two corpora, the BNC for British English and the COCA for American English. The queries are different, because the corpora have different tagsets, so the query for the BNC had to be adjusted somewhat in order to be used with the COCA. The queries are:
(1) —
(2) —
The results are that for the first query I found 435 hits and for the second query 522 hits, which shows that phrasal verbs are used in academic texts in British English and American English. There are slightly more hits in COCA than in the BNC, but since the COCA is a lot bigger, this may not be a real difference in use.

version 1

Phrasal verbs do occur in academic texts, both in British and in American English; in the BNC 5% of all verb forms are phrasal and 7% in the COCA.

version 2

Version 1 is wordy and somewhat clumsy; it contains a lot of redundant information (tagsets, corpus sizes, query syntax, absolute frequencies) and redundant linguistic expressions (“the results are that”) and takes forever to come to the point. Version 2, by contrast, gets the preliminaries out of the way in as few words as possible so as to be able to move on to more interesting things fast. The relevant information is there nonetheless. If you present relative frequencies (percentages), the reader knows that (a) you’ve consulted a corpus, that (b) you’ve run different queries, and (c) that you’ve done your math, taking differing corpus sizes into account. FU-specific technicalities should not be in the paper, they are irrelevant to researchers working with other corpora or different software. In a nutshell: Tell us what you’ve done, but not how you managed to do it.

As you gain experience, condensing information will come naturally. Consider the next example:

Due to the length of this term paper, I will not be able to present the findings of my investigation of all phrasal verbs, so only the use of point out and carry out will be compared.

version 1

For reasons of space the analysis will be restricted to point out and carry out.

version 2

For reasons of space X will… is a set phrase in academic literature. No need to lose more words about saving space than this. Once you read more papers in linguistics, you’ll automatically start using these convenient little chunks yourselves. We all know that academic papers have word limits. Restricted to entails that there were more things about the topic that could have been mentioned, so there’s no need to spell this out.

Another frequent thing is “In her paper”. Instead of writing “In her paper, Freund argues that the progressive also has subjective or non-aspectual function […]”, why not just say “Freund argues”? We know that it must have been in a paper, and you’re going to have a citation anyway. Or, if you agree with what is being argued for, just say “The progressive also has non-aspectual functions (Freund YYYY:pp)”.

You can also, of course, save space by leaving things out entirely. Remember to refer to your research question. What does your reader need to know in order to understand your main point? Everything else needs to go, unfortunartely. Typical things that you should consider deleting entirely include:

• pre-introductions
  Try to come to the point fast. The introduction is for explaining what you are going to do in your paper and (maybe) why, but you do not have to explain that linguistics is fascinating, that English is a world language, that languages change all the time or that speakers do not think about language the way linguists do. These things are commonplaces that waste space and test the patience of your (busy) reader without furthering your argumentation. Write the introduction after everything else has been written; then, once you’re done, check whether the first paragraph can perhaps be deleted without anything important being lost. (I know what I’m talking about, being a notorious pre-first-paragraph writer myself, constantly tempted to put things into ever larger contexts.)
• yourself
  Don’t preface your study with “As a prospective teacher, I really need to have a perfect command of English grammar, and that is why…” or something to this effect. Your own struggles as a learner are not at issue, and you should not diminish the general significance of your study by making it appear to be solely a learner’s issue. You and your readership are interested in linguistics. For its own sake. There’s no need for an apologia to get you started.
• definitions
  If you really need to define things, be sure to find a linguistic definition, not a Duden or Brockhaus definition, and if you have a definition you find useful, don’t add more definitions that are less useful (unless you’re writing about the history of scholarship or you wish to discuss the various merits and shortcomings of different perspectives the phenomenon under study). If you’re writing about something that can safely be considered basic knowledge in linguistics (i.e. if it’s something that your fellow students in another seminar or module can be expected to know), you do not need to quote a definition.
• theory
  Some argumentations require a solid theoretical foundation, but you need to know what you need it for. Don’t re-narrate theoretical issues you never end up referring back to in the rest of the paper. You get no points for correct but irrelevant pieces of knowledge, even if they are on the same general topic. If it doesn’t help answer your research question, it’s irrelevant. If you feel that what others have called a “theory” is common sense, you can usually be brief about it.
• historical background
  (See separate post.)

If you’ve left out everything unnecessary and formulated everything else concisely and you still cannot cover your topic in 2000 words, you may need to re-think the topic or save it for your bachelor’s or master’s thesis.

So-called stylistic errors are really about logical connections. People have all kinds of opinions on what is and is not good style (appropriate style for a given purpose would be a better aim), but where things get plain wrong, it can’t be good style. I’d like to share a passage from A Table Alphabeticall… with you:

Svch as by their place and calling, (but especially Preachers) as haue occasion to speak publiquely before the ignorant people, are to bee admonished, that they neuer affect any strange ynckhorne termes, but labour to speake so as is commonly receiued, and so as the most ignorant may well vnderstand them: neyther seeking to be ouer fine or curious, nor yet liuing ouer carelesse, vsing their speech, as most men doe, & ordering their wits, as the fewest haue done.
 (A Table Alphabeticall, conteyning and teaching the true writing, and vnderstanding of hard vsuall English wordes, borrowed from the Hebrew, Greeke, Latine, or French. &c. 1604)

This is something we should all pin to the walls above our writing desks. Organize your thoughts exceptionally well before you start, then write in ordinary language.

Many of us, however, produce nonsense. That’s nothing to be ashamed about, we have a right to produce nonsense every now and then. But you shouldn’t let your nonsense survive the proofreading stage. Here are some examples:

1. Primarily, address forms became significantly longer. Whereas it consisted of two words at first it then changed into ten within 80 years.
2. I argue that the origin of polysemy is due to using metaphor and metonym which is the most common cause for using it.
3. Between 1900 and 1949 the OED lists 29 verbs related to drug use.
4. Des Weiteren wird die These der verbundenen mentalen Lexika durch die Annahme begründet, dass Lexikoneinträge verschiedener Sprachen miteinander verbunden sind…
5. Konkretisiert man dies, stößt man auf einen weiteren, noch sehr abstrakten Begriff […]. Hierbei handelt es sich um Perspektiven […].
6. Es konnte gezeigt werden, dass selbstorganisiertes Arbeiten […] positive Auswirkungen […] haben dürfte.

You get the point, I hope. Here are two general hints:

1. Watch out for the word it. Does it refer back to the last neuter singular NP in your text?
2. In linguistics, make sure your words mean what you want them to mean. Did address forms really change into ten words? No. The length of address forms increased. Even that is not really true. The average length of the address forms in the letters included in the Corpus of Early English Correspondence increases from two to ten over a span of eighty years, according to this term paper. If that’s what you mean, say it like that. It’s not too technical. My students of German as a Foreign Language used to write things like Der Fernseher ist gesunken when they meant that The number of households with a TV-set decreased. I’m afraid there’s no simpler way of putting it. We need to be this precise.

The website of the Plain English Campaign says:

Imagine you are talking to your reader. Write sincerely, personally, in a style that is suitable and with the right tone of voice.

When you are talking to your reader, say exactly what you mean, using the simplest words that fit.

Your term paper is not going to be written in Plain English, of course. We want to see idiomatic academic English. But within that, “write personally” and “say exactly what you mean”. Read your term paper to a friend. If it’s too embarrassing, re-write in plain style. If your friend doesn’t understand a word of what you’re saying, explain it to them in your own words. And then use those words in writing as well.

I always explain this in the introduction to linguistics, and I always explain this again when I talk about term papers in my seminars, and it’s on the PDF on term papers as well, but somehow, people either seem to get confused or don’t find this important, so I’ll address the issue here as well: when to use single quotation marks and when to use italics.

When you are referring to the form of a phrase, word or morpheme, you should use italics. When you want to talk about the meaning of linguistic material, you should use single quotation marks:

1. The Old English word stræte goes back to Latin strata, which means ‘straight (road)’.
2. Beoncé introduced slay in her lyrics, as a synonym for succeed.
3. Beoncé introduced slay in her lyrics, in the sense of ‘succeed’.

Double quotation marks are reserved for direct quotations. Don’t use them for anything else.

Slashes are used for phonemic transcriptions (and brackets for phonetic transcriptions). Angle brackets are used to mark letters. Curly brackets are sometimes used for morphemes, but morphemes can also be put in italics (with a hyphen), when their form is meant:

4. The German infinitive ending {en}, though spelled with an <e>, is usually pronounced [n̩], [ŋ̩] or [m̩].
5. The German infinitive ending -en, though spelled with an <e>, is usually pronounced [n̩], [ŋ̩] or [m̩].

Short answer: Why on earth shouldn’t you?

Long answer:

Some of you may remember in-class discussions about the English progressive construction, where I argued, in a nutshell, that it is silly to assume that you can teach people when to use progressives and when not to based on so-called triggers like yesterday or usually or the type of verb. She was loving every minute of it is perfect English. What you need to do is ask yourself what the function of the progressive construction is, describe it accurately, and then teach learners the form along with its function(s). In language teaching, mechanistic shortcuts that ignore function just do not work in the long run, and they don’t really make anyone’s life easier, whatever people’s hopes may be.

The same applies to passives and the pronoun I. Don’t ask whether you are allowed to use passives in this or that genre with this or that lecturer, ask instead what a passive construction does and whether you want it. Don’t ask whether a specific pronoun is banned or not, think about its functions. Apparently, there are people who teach students that I should be avoided in general, but this is one of those mechanistic shortcuts that does not, in the end, teach people the real point. In a sentence like (1), I is perfectly fine; in fact, anything else (e.g. a passive) would be misleading, since it is vital to know that it is the author’s practice that is being described, not general practice. What we don’t want, and this is probably what people mean when they say “don’t use I”, is information as in (2), which is irrelevant to an academic readership.

1. I use the term indirect object to refer to the second non-PP verbal argument in linear order, in opposition to oblique object, by which I mean an object in the shape of a prepositional phrase. (constructed example, based on a true story 🙂 )
2. I decided to write this paper on Indian English because, when I was five years old, my parents and I spent three years in… (from a term paper, adapted)

(This takes us to another misconception. When essay writing coaches say that it is a good idea to state why you find a specific topic interesting, what they mean (I hope) is that it can’t hurt to explain the significance of your findings or conclusions to the linguistic community. What they hopefully didn’t mean to suggest is that your lecturer wants to read about how you personally came to study English. No offence, but you yourself are just not relevant. 🙂 Nor do you have space to waste.)

Back on topic. What is the function of passive constructions? Passive constructions topicalise the patient by omitting the agent or introducing them later in the clause, in which case the agent is in focus. There can be various reasons for wanting to do this. In (3) below, a scientific concept is introduced by a reference to a seminal paper. Mentioning the subject matter first and the author at the end of the clause makes it possible to link the clause to the previous discourse (this is about case grammar) and to what follows (the author is taken up as the topic of the ensuing relative clause). Passive constructions like the one in (4) should indeed be avoided, because they are ambiguous with respect to who the agent is (in this case, whether the author is using the OED’s semantic categories or her own).

3. Case grammar is a linguistic theory that stresses the importance of semantic roles in an effort to make explicit the basic meaning relationships in a sentence. Case grammar was developed in the 1960s by American linguist Charles J. Fillmore, who viewed it as a “substantive modification to the theory of transformational grammar” (“The Case for Case,” 1968). (Nordquist, Richard. 2016. “case grammar”. https://www.thoughtco.com/case-grammar-linguistic-theory-1689744, accessed 26 June 2017)
4. The words are categorized semantically. (from a term paper, adapted)

So, long story short: Please decide for yourselves.

This is a hyphen: -
We use it in compounds like faint-hearted or pre-recorded. In German it is called Bindestrich or Viertelgeviertstrich.

This is an n-dash: –
It’s so called because it has the length of the letter n. In German they call it Halbgeviertstrich or simply Gedankenstrich, and it is used together with spaces to mark breaks in sentences. Wikipedia says (under “Halbgeviertstrich”):

In Appositionen, bei Parenthesen und erklärenden Einschüben – etwa in diesem Beispiel – kann der Gedankenstrich das Komma oder die Klammer als Satzzeichen ersetzen.

In English, the n-dash indicates spans such as time or page ranges, as in this example (from Wikipedia, under “n-dash”):

The French and Indian War (1754–1763) was fought in western Pennsylvania and along the present US-Canada border (Edwards, pp. 81–101).

This is an m-dash: —
It’s so called because it has the length of the letter m. In German, they call it Geviertstrich, but it is rarely used. In English, the m-dash is regularly used, without spaces, where the n-dash (with spaces) is used in German:

the fire drill—it was chaos

Got it?

Im Englischen kann man, ebenso wie im Deutschen, attributive Adjektive, auch wenn sie Partizipien sind, mit Adverbien modifizieren:


# Query: BNC; [pos="AT0"][pos="AV0"][pos="VVN"][pos="N.*"]::match.text_genre="W:ac.*";
#---------------------------------------------------------------------------
72543771: to play . ( n contrast both to [[[ the genetically based view ]]] of literary
58651450: hristians under Ottoman rule . [[[ The newly enthroned Tsar ]]] Nicholas I
26334781: ill and knowledge . He accepts [[[ the widely held view ]]] that the
89949649: ty . Alveolar lymphocytosis is [[[ a well recognised feature ]]] of other
18753338: o finding possible referents , [[[ a fully resolved LF ]]] which is
54231140: sed the opinion , supported by [[[ a carefully researched argument ]]] , that the
59179205: ype III domain encoded between [[[ the previously identified type ]]] III repeats
38783431: lf when I went as a visitor to [[[ a beautifully equipped laboratory ]]] to work
18012693: ndardized interview as used in [[[ a carefully conducted sample ]]] survey is
43300094: e established that captopril , [[[ a well known angiotensin ]]] converting enzyme
55653059: ned before . Although there is [[[ a well developed theory ]]] of how
88049224: nancy Abstract Coagulopathy is [[[ a well recognised complication ]]] of peritoneovenous
[...]


Das funktioniert aber nicht mit allen Partizipien gleich gut, und die Adverbien sind meist Intensifiers oder Downtoners (highly, deeply, carefully, …), seltener Raum-/Zeitadverbien wie newly oder previously. Präpositionaladverbien kommen so gut wie nicht vor; eine Suche nach already, above, before, afore und earlier im BNC ergibt nur drei Treffer:


# Query: BNC; [pos="AT0"][word="already|above|before|afore|earlier"][pos="VVN"][pos="N.*"]::match.text_genre="W:ac.*";
#---------------------------------------------------------------------------
38665095: . Finally , the crisis killed [[[ the already enfeebled Dreikaiserbund ]]] . The
59288442: howing that point mutations in [[[ the afore mentioned elements ]]] resulted in
85176940: introduced to be " added " to [[[ an already identified subject ]]] thus has


Die Möglichkeiten sind sehr viel eingeschränkter als im Deutschen. Konstruktionen wie "the later analyzed synonyms" oder "the below quoted example" sind äußerst unidiomatisch; "an already identified subject" geht, "the already identified subject" ist schon deutlich schlechter. Natürlich gibt es das schöne Adjektiv aforementioned, aber es ist eben ein (univerbiertes) Adjektiv, keine Phrase.

Ich warte noch darauf, dass jemand eine (sprachvergleichende?) Hausarbeit zu diesem Muster schreibt. In der Zwischenzeit gehen Sie bitte auf Nummer sicher und verwenden Sie nachgestellte Adjektive:


# Size: 291 intervals/matches
[...]
# Query: BNC; [pos="AT0"][][word="mentioned|quoted|cited|discussed"][word="already|above|before|afore|earlier"];
#---------------------------------------------------------------------------
56000166: ’ . This provision overcomes [[[ the difficulty mentioned above ]]] that where
24796596: indeed been sent to prison for [[[ the offence mentioned above ]]] , and my
58998124: several hospitals . In view of [[[ the factors discussed above ]]] it is most
19208498: ze and how active you are , so [[[ the figures quoted above ]]] should be
104541276: ng verse ) . 14.2 Assimilation [[[ The device mentioned earlier ]]] that produces
1334677: uitively . John Bayley , after [[[ the remark quoted above ]]] , goes on to
104527184: e the syllable it relates to ; [[[ the words quoted above ]]] will thus
70690852: arate species , except perhaps [[[ the features mentioned above ]]] but this
18756817: to suffer from inaccuracies of [[[ the type discussed above ]]] . Again , we
22070827: o line , let us review some of [[[ the factors mentioned earlier ]]] . There is
96961433: has been largely superseded by [[[ the drugs mentioned above ]]] . For maximum
46766802: hich minister to that public , [[[ the works cited above ]]] might never
[...]


I’m no expert on official punctuation rules, and there’s quite a bit of freedom in English anyway, but here’s one thing that will reduce mistake (and annoyance) rates drastically for some people: Germans, please don’t transfer German Kommaregeln to English! I know you’ve probably put a deal of effort into learning this in high school, but German punctuation rules, I regret to inform you, will lose rather than win you points in English. So please: DON’T PUT A COMMA IN FRONT OF EVERY THAT! That can be a relative pronoun (like which, which requires a comma if/because a non-defining relative clause ensues) or a conjunction, and in neither case does it need a comma!

Personally, I don’t think we should transfer English punctuation practices to German, either. I never put commas in front of the main verb (as in Anders gesagt, sind…)—after all, German’s a V2 language, and the position before the verb can be filled by complex constituents, so there’s no reason why I should. 🙂 But I think the Duden’s against me on this one. 🙁 In any case, you should have a good reason for putting a comma before the main verb in German. This is a general rule of thumb, actually: You should have a good reason for any comma you put.