Stylistic analysis of Lengsel’s “Tales of Lost Love” lyrics (2006)

In 2015 I gave a class called “Linguistic Stylistics” at Potsdam University, in which students had to work on their own stylistic analysis of an English text of their choice. While creating a little “how to plan your presentations” handout, I decided to include a little model stylistic analysis of my own, and having read some of the strange official lyrics to Lengsel’s album The Kiss, the Hope (2006) a few days earlier, I decided to use the lyrics to the song “Tales of Lost Love” as a model.

I am posting this here because I figure someone out there might enjoy it – though it is targeted more at students wanting to know about stylistic analyses than at my fellow Lengsel fans. But here goes:

Below I will give you an example of how I would begin to analyze a short text: These are the lyrics to a song by the Norwegian avant-garde metal / post rock band Lengsel, as apparently issued by the band (most of these lyrics are not actually sung in the song, but other lyrics are, which makes this a very interesting text type in itself, but that is a different story… Let’s take the text as it is):

Lengsel, “Tales of Lost Love” (2006)

(written by Lengsel; lyrics found at

tales of lost love
lost, not because it was bound to,
not lust mistaken for love,
not a fake image that faded
nor helpless attempts to project
vague, inner pictures of love
and its possible potential

but true tales of lost, true love
that slipped through someones fingers,
though not by accident,
as the fingers were filled with regret
from the moment they started to spread
such tales
make me wanna knock myself unconscious

First, what kind of an approach am I going for? This is a relatively short text, so any quantitative analysis would probably not show me any frequency effects or anything I would not see for myself through a ‘qualitative’ look at the text. Nevertheless, just to demonstrate this as a possible starting point, here is the frequency list of all forms occurring more than once in the text:

  •  Number         Frequency     Form
  •   1                    4                      love
  •   2                    4                      not
  •   3                    3                      lost
  •   4                    3                      of
  •   5                    3                      tales
  •   6                    3                      to
  •   7                    2                      fingers
  •   8                    2                      that
  •   9                    2                      the
  • 10                    2                      true
  • Total No. of Word Types: 57
  • Total No. of Word Tokens: 75

The text’s type/token ratio is pretty high (0.76; the analysis you are reading has a type/token ratio of only about 0.32), which means that it has a relatively rich vocabulary and relatively few repetitions of forms. However, this ratio is pretty meaningless when dealing with such a short text (only 75 tokens!). Type/token ratio measurements make a lot more sense when you are comparing longer texts, e.g. different chapters of the same novel, or different novels, etc.

Grammatical words or “function words” are at the top of any word frequency list (Lindquist 2009: 27): If we leave them out, our frequency list contains the items love, not, lost, tales, fingers, true. Interestingly, almost all these words are featured in the first line of the second verse – “but true tales of lost, true love”. This line can be said to encapsulate the topic, the “aboutness” of the poem like no other line.

 Now, here is a brief look at all of the formal levels of language description:

Phonology: The text displays some regularity of stresses, but no strict meter. Each line has about three to four stressed syllables. The only exception is the penultimate line, which is noticeably short and therefore deviant: “such tales”. In this line, both words seem stressable, which seems to “slow down” the pace.

Although the meter is not strict, there is a tendency towards anapest (xx/) in many lines, e.g. “nor help-less at-témpts to pro-jéct”; “as the fín-gers were fílled with re-grét”. That is, most stresses are separated by about two unstressed syllables. This tendency is completely reversed after the foregrounded short (and “slow”) penultimate line “such tales”: The last line is completely trochaic (/x): “Máke me wán-na knóck my-sélf un-cón-scious”. It is additionally deviant in that this is the longest line (five stresses as opposed to the usual three or four).

The text displays some further parallelisms on the phonological level, e.g. alliterations in many lines (lost – love, lust – love, fake – faded, possible – potential, true – tales, slipped – someone, fingers – filled, started – spread, make – me [– myself]). [The word myself doesn’t really count because the /m/ does not belong to a stressed syllable…]

The last line (“make me wanna knock myself unconscious”) shows consonance in the repeated use of nasal sounds (/m, n/).

Graphemics: The text is written in lower-case letters only, and contains what seems to be one deviant spelling (a.k.a. a mistake): “someones”. I would not place too much emphasis on this level, though, as this text is not taken from any printed edition, but this text has been apparently been submitted by the band in a digital format. “someones” is most probably not more than a ‘typo’.

Morphology (and some semantics): The text contains many verbs that are either past tense or past participle forms (lost, bound, mistaken, etc.) and only one verb in the present tense: make in the last line.

Lexicology (and some semantics): Deviation: The adjectives in the text move from falseness and indeterminacy (fake, helpless, vague, inner, possible) to the one semantically opposite adjective true that is repeated twice at the beginning of the second verse. Parallelism: The first line of verse 2 repeats the first line of verse 1, but with the word true inserted:

  • Verse 1, line 1:                tales of lost         love
  • Verse 2, line 1: [but] true tales of lost, true love

Semantics: There is clearly a contrast between the first and second verse. The first verse is filled with negativity (cf. Jeffries and McIntyre 2010: 71-72): Cf. the particles not … not … not … nor, which are balanced by the first word of verse 2: but (as well as by the positive word true, see above). The third line in the second verse is somewhat of an exception, though the not in this line refers to an embedded clause.

One could also look for the semantic “agents” of verb phrases: These are generally impossible to find! Who is it that is losing, mistaking, attempting, etc. in the first verse? It seems the agents are avoided through passive constructions and nominalizations (in the case of attempts). By stark contrast, we do have some subject-verb relationships in the second verse: Here it is “love” that “slipped through someones fingers” and the “fingers” that “started to spread”. It seems the difference between the “tales” of the first and the second verses is that those of the second verse do include active ‘agents’, even if these appear somewhat hazy. The general difference between the tales described (and discarded) in the first verse and those described in the second, is that in the second verse some conscious human action (referred to associatively by the metonymy “fingers”, and paradoxically paired with simultaneous “regret”) leads to the ‘loss of love’.

Syntax: I think this is the most conspicuous finding of all: Syntactically, almost the entire text is one huge subject, and the last line is the verb phrase that finally completes the grammatical sentence! The three repetitions of the word tales structure this very complex subject, so that the whole text boils down to one sentence, which could be formalized as follows:

TALES, not TALESw, but TALESr, TALES make me wanna knock myself unconscious.

 in which TALES represents the neutral first line (“tales of lost love”) as well as its semantic recapitulation in the penultimate line (“such tales”), TALESw represents the ‘wrong’ kind of tales talked about in the first verse, and TALESr represents the ‘right’ kind of tales in the second verse.

This syntactic deviation of the last line (i.e. the fact that it is not simply another noun phrase constructed around the word “tales” like the other lines, but an actual verb phrase that completes the sentence) further contributes to this line’s stylistic foregrounding.

Pragmatics and discourse analysis: The text consists of one grammatical sentence without context, and second-person pronouns are completely absent, which makes it impossible to analyze in terms of conversation analysis, relative speaker power, etc. The entire text is one ‘turn’ / utterance / monolog given without any context. Nevertheless, the speaker makes himself[1] explicit via first-person pronouns in the last line (“make me wanna knock myself unconscious”), which gives the text a certain perspective, or a subjective ‘twist’. The very fact that the subject is so long (essentially an elaboration of the phrase “tales of lost love”) may prompt us to assume that Grice’s maxim of quantity (“Do not make your contribution more informative than is required …” – cf. Yule 1996: 37) is flouted here, IF we could be sure that the utterance was part of a larger conversation (which we cannot – the whole ‘conversation’ may be taking place in the speaker’s mind, for all we know).

Now that we have gone through all levels of language description, the one thing that really stands out is that the last two lines, and especially the very last line, are foregrounded on a large number of levels. So what do we make of this?

From a “reader response” angle, we can say that the verb phrase at the very end (after the 68-word subject!) comes as a surprise to the reader. The text up to this point has been read as a “set of noun phrases in apposition” (cf. Jeffries and McIntyre 2010: 54), to be para­phrased as, “TALES, not TALESw, but TALESr, …”. It is only at the very end of the text that the reader notices that the speaker has actually begun a sentence of which TALES is the subject.

A possible effect of this could be that the reader has been forced to deal with all of the attributes of the ‘wrong’ (i.e. negated) and ‘right’ kinds of tales, so that the reader might have seen the advantages of the ‘right’ kind of tales and cannot help but agree when the speaker finally continues by stating his typical emotional reaction to the ‘right’ kind of tales.

The last line itself – in which the speaker reveals his emotional reaction – is deviant on many levels, as we have seen. Semantically, it is the only really ‘personalized’ line. But what can we say (stylistically) about the wording of this line? Is the phrase “X makes me want to knock myself unconscious” externally deviant, i.e. does it deviate from language use in general?

In order to answer this rather special question, I will turn to quantitative stylistics once more, and search for the phrase “MAKE me want to” in the Corpus of Contemporary American English (COCA, cf., with MAKE representing the lemma MAKE, i.e. all forms of the word, e.g. make, makes, made, making, etc. (The search node “MAKE me wan na” yielded practically no results.)

Which kinds of verbs usually follow “makes me want to” in contemporary English? – A search for any verb lemmas ([v*]) following the string “MAKE me want to” (search horizon set to up to 5 items to the right) in the COCA yields the following results (only the first 25 results are given in the following poorly designed table):

  • Number          Lemma           Frequency
  •   1                    GO                   48
  •   2                    BE                    44
  •   3                    CRY                 42
  •   4                    DO                   32
  •   5                    THROW           25
  •   6                    GET                 24
  •   7                    PUKE               17
  •   8                    SAY                 13
  •   9                    SEE                 12
  • 10                    LAUGH             11
  • 11                    LIVE                 11
  • 12                    SCREAM          11
  • 13                    VOMIT              11
  • 14                    GIVE                  9
  • 15                    KNOW               9
  • 16                    RUN                   9
  • 17                    TAKE                 9
  • 18                    WRITE               9
  • 19                    HAVE                 8
  • 20                    COME               8
  • 21                    STOP                8
  • 22                    WEEP                8
  • 23                    JUMP                 7
  • 24                    REACH              7
  • 25                    SING                  7

Apart from a few verbs that are very frequent in any context and do not carry a lot of semantic meaning at all (GO, BE, DO, GET, SAY, GIVE, TAKE), the list is topped by verbs that denote a negative (physical) reaction (CRY, PUKE, SCREAM, VOMIT, WEEP; possibly RUN and STOP) and verbs that denote violence (THROW, and again, SCREAM). It is conspicuous that the choice we get in the text, “knock myself unconscious”, denotes self-aggression, which is something akin to a combination of these two categories. [A third category is that of positive (physical) reactions (LAUGH, JUMP, SING).]

Thus, the last line so far does not seem to be particularly deviant from ordinary language, apart from the fact that the verb KNOCK is not very high up on the list of usual “MAKE me want to” collocates. The COCA actually does list two occurrences of “MAKE me want to knock” (both from fiction), which are given below.

  • Year:  Concordance line:
  • 2010    …hair thinned down to wisps on top, and a self-satisfied grin that made me want to knock his teeth in.
  • 1996    …but the way he was looking at me made me want to knock on his head and ask if anyone was home.

In both cases, the string denotes a violent or intrusive action against another person (as opposed to oneself). In this respect, the last line could be said to be deviant after all, but more data would be needed to corroborate this.

Another thing to find out would be what nouns / subjects usually precede “MAKE me wanna [negative action]”. I suspect that “tales” would not be among the top words of such a list. Just judging from the two examples above, where the subjects are “a self-satisfied grin” and “the way he was looking at me”, I would suspect that the subjects are usually other human beings (who are behaving in a way that provokes the speaker’s negative reaction), and certainly not “tales”. Thus, the speaker might be said to be treating “tales” (or the characters therein) as if they were human beings; the speaker is describing his reactions to the contents of certain “tales of lost love” in terms that one would normally apply to reactions to encounters with human beings in real life. This is what makes this text a meta-text of sorts, i.e. a “text about texts”. An (admittedly crude) interpretation of the text’s ‘meaning’ would be as a creative and reader-engaging way of saying, “stories about the end of a love relationship, in which the characters are somehow actively engaged in ending their (otherwise healthy) relationship, really get to me”. But, of course, the best way to destroy a poem is to try and paraphrase it.

This has really only been an exercise in order to give you a model analysis for your presentations. I have written more than 2000 words filling six pages, and reading it all out would take about 20 minutes! In other words, this has already been too long! If I were to turn this into a “final version”, I would certainly leave out many things: For one thing, the primary text should not be included in your “written versions” at all. In my analysis, the graphemic and morphological levels, for example, have yielded nothing substantial. The quantitative exercises were included only in order to demonstrate what COULD be done. I would concentrate much more on what is happening on the syntactic level and perhaps mention in passing the other things that contribute to the foregrounding of the last line(s), like the rhythmic shift (from unstressed-stressed to stressed-unstressed).


  • Jeffries, Lesley, and Dan McIntyre (2010). Stylistics. Cambridge Textbooks in Linguistics. Cambridge: CUP.
  • Lindquist, Hans (2009). Corpus Linguistics and the Description of English. Edinburgh: Edinburgh UP.
  • Yule, George (1996). Pragmatics. Oxford: OUP.


[1]     Living in the 21st century compels me to add the footnote that the masculine form here is to be understood as gender-neutral.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: