Conference papers

Here you will find a selection of conference papers in a combination of my text and PowerPoint.

Cultural colonisation: Opus 5 no.3 as a case study.

Paper by Per Dahl at International Grieg Conference in Bergen, Norway 17 – 19 November 2023: Current Musical Heritage.


Edvard Grieg wrote Melodies of the Heart as his opus 5 based on the text of H. C. Andersen in Copenhagen in 1864 and gave it to Nina for Christmas. The third song, Jeg elsker Dig, became very popular. It was first printed in Copenhagen by Chr. E. Horneman, in 1865,[1] later by Wilhelm Hansen from 1879, and Jeg elsker Dig as part of "Romance og Sange" from 1880, no.9 in Volume I. Also, the publisher C. F. Peters in Leipzig, covering the European market except for Scandinavia, published the song in their Grieg-Album of songs, as no.13 (first) in Album II in September 1875, and in a new edition in 1888 where there suddenly was a second stanza in German (by Frank van der Stucken) added to the one stanza poem of H. C. Andersen. In addition, Peters printed the two-stanza version as sheet music in over 60,000 copies in 1883-1906.[3] However, there was no indication of repetition sign in the music.

In this paper, I will comment on some of the hierarchy of power and ideas that can be attached to the performance traditions of Grieg's Opus 5 no.3 that can unveil some elements of our musical heritage. The data is from my dissertation's discography of 318 recordings of this song, where I have analysed the performance in 214 of the recordings. I have used the expression 'cultural colonisation' in my paper's title, implying that power hierarchies govern the development of art expressions and artefacts in different cultures. Opus 5 no.3 is closely dependent on a cultural context to be created as a composition and as a performance. In both situations, hierarchies of power influence the result. Therefore, I find it relevant to draw some parallels to colonisation, defined as a process of establishing foreign control over target territories (here: the musical work) or peoples for the purpose of cultivation, often by establishing colonies (here: musical style) and possibly by settling them (canonisation of works/composers/performers). Especially in classical music, the development of different styles, the hierarchy of works, performers, composers, and narratives of music history parallels establishing colonies and determining the discourse of music communication.

The text

Let's start with the text. Grieg used only four of Andersen's eight poems in Hjertets Melodier from 1830, when Grieg composed his opus 5. Grieg reorganised them, making the story more convincing as a declaration of love than Andersen's preference for distant admiration. Even after repeating the two last verse lines (Jeg elsker Dig, i Tid og Evighed), opus 5 no.3 is among the shortest songs in the classical repertoire. Adding the extra stanza in German in the Peters editions can be seen as a necessary adaption of this song to the standard of a German Lied in size and content. While the original text by Andersen is intensely focused on how falling in love makes the I-person's hierarchies of thoughts and associations being concentrated towards the Chosen One, the supplementary second stanza relates to the object of love and adds God and Fate into the realm of this song. These elements are not in Andersen's text, but they give associations towards German romantic philosophy and literature. As such, I will call this a cultural colonisation of Opus 5 no.3.

This colonisation becomes even more evident as all the translations of this song to 22 different languages all use the German two-stanza version, not Andersen's original, as a point of departure. There were several (nine) translations to English, all close to the German text until Wilhelm H. Halvorsen's translation in Grieg Gesamt Ausgabe (Peters), which is based on Andersen's one stanza. It illustrates the conflict between periphery and centrum, i.e., Nordic tone versus German Scholarship in defining musical style. So strong was this cultural power hierarchy that Scandinavian singers singing in Norwegian/Danish up to the 1980s sang Andersen's text twice to make the song fit into the German Lied tradition!

There are, though, two significant deviations from this. However, both reveal the impact of context in presenting opus 5 no. 3. outside of the classical domain, they supplement our musical heritage and might change the common understanding of what Jeg elsker Dig! is all about. The text in the musical Song of Norway by Robert Wright (still two stanzas) has the more updated title 'I love You' instead of the traditional translation 'I love Thee'. The declaration of love is much more focused on the beloved's body, face, lips, and eyes, and there is nothing left of the delicate balance between the sensual 'Jeg elsker Dig'and the morally binding 'I Tid og Evighed!' The God and Fate from the German text is replaced with the Northern Star and the world's relentlessly growing.

A completely new German text came with the recording of Julia Migenes in 1987 by Gerhard Bronner. It has much of the same allusions to bodily references to love. Still, the most striking change is an adaption to the selfish world in the Yap time: Instead of "Ich liebe Dich in Zeit und Ewigkeit", the moral obligation is reduced to "Ich liebe Dich so lang' Ich fühlen kann". I call that an example of the cultural colonisation of human love based on the commercial market's use and discard-mentality.

Indicators of cultural colonisation in the recordings of opus 5. no. 3

The modern assumption that interpretations of old recordings had a lower tempo was not confirmed in my analysis of 214 recordings made from 1899-2000. The second stanza was usually performed slower, but only compared to the same singers performing the first.

However, the distribution of languages in the recordings gives exciting indicators of the hierarchy of power and cultural development in the recordings of this song. The following Figure shows the distribution of recordings in the main languages in the market. Up to the mid-1930s, the market was dominated by recordings in German. Then came a period where recordings in English dominated during World War II. In the afterwar period, no clear dominance could be detected until the authenticity movement, seen as a branch of Early Music, made Norwegian/Danish the preferred language for singers of any nationality when recording this song.

There is also an exciting development in the distribution of voices. In contrast to the other very popular Grieg-song, Solveigs Song, recorded only by female singers, Jeg elsker Dig can be sung by both male and female singers. It has mostly been recorded by high voices (soprano/tenor). However, after 1955, the number of tenors recording this song decreased dramatically, and the sopranos dominated the market. The song's theme, love, should be relevant for all humans. Still, this development might be affiliated with the very divided roles in the family in the afterwar period, with the Father working to get money and the mother caring for love.

Grieg wrote only a piano accompaniment for this song, and in the acoustic era, the piano accompaniment was preferred. After Max Reger produced an orchestra score to Opus 5 no. 3, several recordings were made, and in the electrical era (from 1925), there were more recordings with orchestra than with piano. The piano was long established in the home of those who liked to play music, but to have an orchestra playing at home was one of the great fortunes of the gramophone. This cultural reorientation was interrupted by the war, and like the choice of language, an apparent new tendency developed from the 1970ies where this song was to be accompanied by piano again.

There is an interesting development in the tempo deviations in the recordings. I divided the song into sections of text lines and pure accompaniment (sections 2 and 4 in the next PP). I calculated the mean tempo in each section for the four epochs of the recording industry. The tempo in the first text line in each recording was set to the norm of the recording, and tempi in other sections were calculated to percentage-wise deviations. There is a big contrast between the acoustic era pianist, who seems eager to get through the song playing much faster in his interludes. While in the digital period, the pianist plays slower than the singer. Now, the pianist is not only accompanying but also taking part in interpreting the song and its text. This change of power between singer and pianist in recording a Grieg song might be something Grieg anticipated, writing in his diary about a concert in Prague where he accompanied Cally Monrad: "Cally Monrad sang ravishing, beautiful, but far from my intentions."

Even though the duration of performing this song did not change during the whole century, there was a dramatic change in the performance practice in the 1950-60ies. That is when vinyl recordings replaced the 78 records, and studio recordings were mastered on tape for editing before the production of the gramophone record. In the growing economy, the distribution of records reached new markets. The change in performance practice could be seen in my data when I calculated the Standard deviations for each section in all recordings and grouped them according to the four epochs in the recording industry. The Figure in shows that the standard deviation decreased significantly around 1955, indicating that independent of tempo in the performance, the tempo deviations in each section become much less from 1955 than before. There was no change in the vocal or piano equipment then so the difference might be connected to a broader distribution of gramophone recordings to music conservatories: music students and musicians developing their musical taste more commonly used those records. Listening to records substantiated an understanding of different traditions and historical practices possible without being present at concerts. It enabled the music student to create a hierarchy of performers. At the same time, the record industry could intensify its branding of cultural hegemony, making the gramophone artist the peak artist in music life. The gramophone record was easy to distribute, and companies and cultural departments of Embassies used collections of records given to institutions for the purpose of cultivation. At the same time, this was a legitimation of cultural colonialism.

The last element I will point to is the impact of technological development in the record industry. There are two main turning points: the change from analogue to electrical recording in 1925 and the digital revolution presenting the CD in the 1980ies. The record companies are primarily interested in selling as many records as possible, and such new technological developments made it essential to include as many listeners as possible. My analysis classified the recordings into two categories: commercial and concert traditions. The Figure in (PP10) shows that the only time the number of commercial recordings rises and the concert tradition is reduced in these two critical periods of technological changes. The extreme number of concert recordings in 1991-95 can be related to the Grieg Jubilee in 1993. As such, it demonstrates the many contributors to branding cultural hegemony.

Conclusion, or a sum up.

Studying the performance practice of Opus 5 no.3 as documented on gramophone records leads to a revaluation of the concept of classical art music as an ahistorical phenomenon. The case study made it possible to point out different kinds of impact on the performance tradition, adjusting our notion of what can belong to the world of Grieg's music.

Three leading hierarchies of power were detected:

  • The publishing of the work. The score has an exceptional position in classical music, so the work is often identified with the printed edition. That makes the publisher a powerful agent in developing our musical heritage. The notation in opus 5 no.3 is inconsistent in the different editions, even those Grieg supervised. However, the translation to German made Opus 5 no. 3 a contributor to the cultural branding of the German Lied.
  • The performer's realm of interpretation. The case study exemplified several changes in the performing practice, tempo adjustments, the role of the accompaniment, the language of the performance, and what kind of voice dominates the gramophone market. The changes in the distribution of languages represented on recordings illustrate a development in the transnational understanding of Grieg's music and an adaption to the German Lied canon. The performance practice constantly evolves depending on the actual cultural hegemony. I used the musical Song of Norway as an example that bends the concert tradition towards focusing on commercial effects.
  • The development of recording techniques. The material was divided into four epochs, and the main changes were from the acoustic to the electrical technology in the 1920s and the introduction of digital recording in the 1980s. The gramophone companies are more interested in market shares than the inherent value system in classical music. This case study shows they are essential to a transnational cultural understanding.

This kind of narrative of music communication significantly impacts our current musical heritage.

So, my punchline will be:

The case study demonstrates how artists constantly try to establish control over the work (target territory) for cultivation by establishing a stylistic expression (artistic colony) and settling it in a solid product, a gramophone record (canonisation of works).

[1] Horneman had published Griegs Poetiske tonebilder for Klaver op.3 og Seks sanger for alt og Klaver op.4, composed in 1863-64 and published in 1864.

[2] Wilhelm Hansen bought Horneman in 1879 and kept using the printing plates (no.91)

[3] Two versions: German/English/French text (1883) and one with Italian/Spanish text (1890)


Jeg elsker Dig! Lytterens argument. Grammofoninnspillinger av Edvard Griegs opus 5 nr.3. 

Dr.philos. avhandling, Universitetet i Stavanger 2006.

Modes of Communication in Narratives

Grieg Research School Bergen June 15, 2022

In this paper, I take for granted that narratives exist only as modes of communication. The question of medium is subordinated to the fact that humans use and develop narratives in language, music, and dance. Common to these modes is the possible connection to an artistic expression which presupposes an opening for a symbolic understanding of the actual impression. That makes the constituency of communication my point of departure in developing a multi-layered model that can detect distinctive modes of communication in narratives. After a presentation of a theoretical platform, I will visit one of Stravinsky's narratives about music.

The objective elements in a narrative are the ontological entries necessary to identify the narrative. However, there is also a cultural and an epistemological dimension in our understanding of a narrative. By expanding Ogden's triangle into a multi-layered model where many relevant concepts for analysing narratives are structured into the three corners of the model, modes of communication in narratives can be distinguished.

All modes of communication consist of signs and expressions in an exchange between humans. The difference between a sign and its expression is fundamental in understanding narratives, and I will describe a familiar situation, handshaking, as a narrative to develop some analytic tools/concepts. (All theories should have a practical fundament).

A person (Arne) approaches another person (Anne), and Arne lifts his right hand to greet the person he approaches. When the approaching Person, Anne, has identified this act as part of a greeting ritual, the person in question will (likely) lift her right hand to return the greeting.

I will describe this situation with terminology across three levels making a conceptual grid for analysing narratives.

Level 1: A1 intends to greet A2 and acts by raising his right hand (element B). (If Arne was walking in a diagonal stride waving his hands, at one point, he decides to leave the right hand in the front position). A2 identifies the act (B) as a product in the context of greeting rituals (element C) and lifts her right hand. Two preconceptions illustrate this situation: 1) To identify a sign is to categorise a segment of reality (the act) as a single element (product) classifiable in a pre-existing category (context). 2) As far as both perceive the act of greeting alike, that is, relate it to the same familiar context of greeting rituals, a meaning transference of A1's intention to A2 has been possible. Even if they/we lack this shared understanding, the sign or gesture can still be identified, but the consequence depends on the interpretation that A2 attributes to the sign. Interpretation is thus more than the identification of signs. Interpretation also involves situating sign identification within the interpreter's mode of expectation (A2's intention), where A2's appraisal of consequences come into play in addition to the identification of the sign and its contexts.

Level 2: A person has a belief (the personal meaning) (A) that is shown through the signs used (B) to express the intended action (C). Person A1 is responsible for the belief and the sign used. His intention includes the wish for the sign to be seen as an expression following his meaning. However, it is up to A2 to interpret the sign and connect it to an expression. The link between the sign and the expression is arbitrary and depends on Anne's mode of expectations. 

Level 3: A person has a horizon of understanding (representational world) (A) evident in the choice of ideological statements (B) in given contexts (C). In our case, the person has a conception of the elements in greeting rituals and situations where they are typically used. Implicitly, the context of these elements is culturally conditioned: what counts as belonging to a greeting ritual varies from culture to culture. In our culture, the right hand is the one for greeting. This understanding presupposes an imaginary organisation of the elements or signs to the ideological frames constituting the cultural circle. This organisation can be called a pre-judgment in Gadamer's sense, as it is not the conscious act of an individual but belongs to her way of being.

The crucial elements in this model are the arbitrary connection B-C and the plural of the person in position A. The model illustrates how communication has two significant challenges: 1) The sender and the receiver can have different horizons of understanding, and 2) The arbitrariness of the sign/symbols of the message makes the communicative meaning of the message dependent on the receiver. Therefore, identifying a sign/symbol will be part of a social construction of reality that makes communication possible.

We can use different methods of explanation (discourses) for the different relations in the model in analysing narratives. The beliefs of the person can be articulated as intentional explanations. The sign is identified through an operational explanation (operationalisation of the sign). Understanding what it expresses can be explained through causal explanation (which is to say, a discovery of the causal connections obtained between expression and the perceivers of the expression). 

Focusing on narratives, I will add 'metaphors', 'allegorical', and 'synonyms' in parallel to the scientific explanations.  Metaphors refer to the relation between the Sign and its Meaning, whereas allegorical expressions should establish a link between the Sign and its Expression/Context. Synonyms are mostly connected to words and literary associations. As narratives and art expressions can establish non-literary meaning, I find it more plausible to use a link to the concept of the extended moment as the third relation. In the extended moment, we place the identification of an impression in an interpretational process using the present, past and future scaffolding the meaning construction.

The cultural dimension of the greeting ritual here described as a social construction of reality could also be labelled as a transcendental object in a phenomenological discourse. To fulfil the model's philosophical aspiration, the three corners include links between 1) the person transmitting the narrative and subjectivity, 2) the message/actions and objectivity, and 3) the narrative as a product in the context of intersubjectivity as the receiver experiences it. With these concept layers, the model unveils a continuum from the more physical/observable elements in the centre to the more philosophical concepts in the outer circle. Being aware of the concepts placed in the different layers is of great help in analysing and arguing the essence of a narrative.

Narratives are often associated with artistic expressions. I have, therefore, added an extra layer to the model making three critical domains for studying narratives. In discussions, Aesthetics, Logics and Ethics might not be expressed, but they are often used to develop arguments on a lower level. It is crucial to remember in what relation they belong.  An aesthetic issue is about the connection between a person's meaning and the use of Signs/Symbols in action/performance, not about the relation between the action and its product, and not at all about the product's relation to the receiver (which is an ethical issue). The Logics deals with the relation between the Sign and its Expression. Quite often, however, aesthetic statements deal with a logic not respecting the arbitrariness between sign and expression and merge the Context and the Ideology. Then an authoritative analysis might end up as an authoritarian analysis. Aesthetic arguments can support the intentionality of the interpretation. However, the performance needs to be operationalised within a chosen functionality system between the Sign/Symbol and the expression. The receiver's response in interpreting the identified product is the first step within the domain of ethics. By contextualising the response, the receiver will react/make an ethical judgement not always as a conscious act but certainly as a part of the epistemological process attending narratives.

I will now present some findings from a case study in my book: Modes of Communication in Stravinsky's Works. Sign and Expression (Routledge 2022). Through his books and interviews, Stravinsky created many narratives about music and how to perform music. The multi-layered model unveils possibilities to reduce the discrepancies and enhance the understanding of his narratives.

A core element in Stravinsky's aesthetic is sometimes reduced to the slogan: "Music is powerless to express anything at all."  He presents the idea in his Autobiography and repeats some of it as a question in his Norton lectures: "Do we not, in truth, ask the impossible of music when we expect it to express feelings, to translate dramatic situations, even to imitate nature?" (PoM p.77) Stravinsky's list of expected expressions here mixes ontological (objective) elements like dramatic situations and nature and epistemological elements like (subjective) feelings. Adding 'in truth' to the statement makes it look like a scientific law, which underlines Stravinsky's search for objectivity in the ideology of absolute music.

A complete presentation of the slogan is to be found in the Autobiography. Here I will present a close reading of the beginning of this quote, the numbering of the elements prepares my complete argumentation in the book. 

For I consider (1) that music is, by its very nature (2), essentially powerless to express (3) anything at all, whether a feeling, an attitude of mind, a psychological mood, a phenomenon of nature, etc..... Expression has never been an inherent property of music (4). That is by no means the purpose of its existence (5). If, as is nearly always the case, music appears to express something (6), this is only an illusion and not a reality (7). It is simply an additional attribute (8) which, by tacit and inveterate agreement (9), we have lent it, thrust upon it, as a label a convention (10)- in short, an aspect unconsciously or by force of habit, we have come to confuse with its essential being (11). (AA pp.53-54)

When Stravinsky uses 'I', it could be the composer talking, but the rest of the text makes the 'I' in (1) being an observer of the common phenomenon called music. Adding a reference to nature (2) reinforces the search for an objective description of the music. The conclusion (3): powerless to express anything is not a necessary outcome of this chain of reasoning. What Stravinsky does here is a rather messy phenomenological reduction pointing to the essence of music as something powerless to express. The problem with this argument is that expression is an epistemological entity, and Stravinsky uses it as if it was an ontological entity in defining music. His examples (a feeling, an attitude of mind, a psychological mood, a phenomenon of nature) result from actions in a human mind, not accessible to an observer. Stravinsky continues to trace music to a natural phenomenon in (4), arguing that expression has never been an inherent property of music. This argument is only valid if the music was detached from human activity. The chain of reasoning is taken even further in (5) when he argues, quasi ontogenetic, that expression by no means is the purpose of music's existence. Most anthropology studies enhance music's ability to create expressions in developing human societies. Stravinsky accepts that music appears to express something (6), but this expression is not part of music; it is only an illusion (7) and not a reality. Again, he keeps the concept of music towards objectivity and dismisses the human sphere of experience. The problem of not accepting expressions as reality pops up in his introduction to his Norton lectures published as Poetics of Music; "it is difficult to talk about music if one considers only its material realities", an example of many contradictions in his reasoning.

In Poetics, he talks as a composer introducing himself and his topics. He is a person with skills and meanings, and his horizon of understanding makes it necessary to open music to more than material realities. In his Autobiography, he is the critical observer trying to define music so that his ideas about absolute music can be promoted. We have here an example of Stravinsky using different modes of communication on the same issue. Some of these narratives are incongruent with the music he created and his interpretations of his musical works when he was the performer/conductor. The multi-layered model unveils possibilities to reduce the discrepancies and enhance the understanding of his narratives.

As my presentation must end soon, I will conclude, hoping that more examples of how the multi-layered communication model can contribute to the analysis and understanding of narratives can come up during the discussion now afterwards. The fundamental mode of communication in narratives is accepting the extended moment making the narrative an epistemological entity based on intentions, identifications, and interpretations. Coming from music, the concept of the extended moment is constitutional to all temporal expressions, including narratives. Therefore, as I see it, analysis of narratives must include the epistemological entities based on both the creator's intentions and identification of signs and expressions and the receiver's identifications of signs and expressions in developing an interpretation of the narrative.

Stravinsky; three periods but two styles?

Birmingham: Music Since 1900. 17-20 June 2022. Per Dahl.

The three-period division of Beethoven's life and works became a model for musicologists presenting the life and works of their composer, hoping to get acceptance for that composer's place in the hall of fame. Sometimes dramatic life changes are connected to works, and sometimes new stylistic elements are taken as evidence for the three-part division.

Traditional analysis has focused on the compositional elements (melody, rhythm, harmony, form) and combined with ideological utterances from Stravinsky and biographical facts, the terminology of his three styles, Russian, Neoclassical, and Serial has emerged. Based on my empirical study of his notation practice, published this year, from which I am to present snapshots here, there is a possibility to rethink this three-part division.

The combination of analysis of Stravinsky's works and his literary utterances is usually based on written material (the scores, manuscripts, his books, and conversations with Robert Craft). My point of departure is that music is a performative practice based on sound, temporal movement, and aural impressions. In my project, I have analysed Stravinsky's notation practice by registering his vocabulary of articulation, dynamic and tempo signs, and literary expressions in the scores. This vocabulary is Stravinsky's direct message to the musicians about how to perform his music. A total of 232 different signs and expressions from 155 published works were put into a database having 5,486 entries. I think this can be an essential supplement to the analysis when defining Stravinsky's style.

From a professional performer's perspective, it is mandatory to be able to produce the tone lengths and pitches in the score/parts. However, the performer needs to know how these fundamental elements of the music are to be articulated and given a musical meaning/character. Stravinsky expresses strong opinions about music performance, interpretation, execution, and musical meaning in his books and interviews. Often it is possible to find incongruencies in Stravinsky's utterances. In my book, I developed a multi-layered communication model, making differentiation of his modes of communication possible. The analysis unveiled Stravinsky as a composer who wants to communicate his ideas of music both as musical works and as literary statements. However, the study shows that he was a better composer than a philosopher of aesthetics, mainly because his understanding of the epistemological dimension in language and music was somewhat blurry.

We have learned from the Early Music Movement/ Historical Informed Performances that style is much more than executing the notated music. Style is primarily a question of the performer's use of conventions in the performance practice. In my study, I have tried to detect some of the conventions in performing Stravinsky's works, in the way he has indicated in the notation. After developing the database, I analysed the empirical findings in a context including Stravinsky's literary statements in his books and interviews and Stravinsky as performer in his gramophone recordings.

This paper will concentrate on the empirical study of his notation practice. After finishing the primary registration, I organised that vocabulary into four different groups with subgroups and classified signs and expressions in further subgroups. The registration was made in a cumulative process. Every new sign or expression was added to a continually expanding list from reading each composition. I did not register the number of occurrences of each sign/expression in each piece.

The four groups of Stravinsky's vocabulary were:

1 Articulation

Universal articulation (sub-subgroups of Staccato, Marcato, Legato, Accents, Two-note slurs, Double articulation), String articulation, Wind articulation, Character articulation

2 Dynamics

Dynamic signs. Cresc. /Dim., Sforzando/sforzato, Dynamic adjustments

3 Tempo

Tempo steady, Tempo adjustments, Tempo character

4 Literary expressions

Cantabile, Dolce, Espressivo, Appasionato, Tranquillo, Piacere, Grazioso, Poco, Energico

The four groups have in total of 20 subgroups. In addition, the subgroup Universal articulation has six sub-subgroups. The total number of signs and expressions in each group is Articulation 51, Dynamics 41, Tempo 69, and Literary expressions 71, which gives 232 different signs and expressions. Depending on the research question, the data was organised in several ways.

I also organised the compositions into nine work categories: Stage works, Chamber, Chamber solo, Vocal and Piano, Vocal and ensemble, A cappella, Orchestra, Piano, and Piano transcriptions. In addition to a distribution of Vocabulary combined with work categories, I also used the traditional three periods for detecting developments in Stravinsky's notation practice. I made the Octet (1923) and The Rakes Progress (1951) the frame for Neoclassical works making the Russian period before and Serial period after these years.

I came to operate with three kinds of the overall description of the notation. Basic notation (instructing how to produce the sound)-related to this: Adjustments (indications of deviations from the primary element/referent defined in the notation, like staccatissimo where staccato has been used). Of quite another kind are the nuances: they are expressive characters where the identification of the sign/symbol has no link to the score, like dolce or poco pesante. All the Literary expressions make references outside the notation, and the same could be said about Character articulation and Tempo character. That makes this kind of information from Stravinsky to the performer related to the performer's idea of stylistic conventions and not being fully decided by the notation chosen by the composer.

I will now present a few of the results from the empirical study making the question of two or three styles the core element. In the total distribution of my material, it looks like this:

As Stravinsky made several revisions of works from the Russian period in the late 1940ies, I compared the original and the revision. I found that there was no significant difference in the use of vocabulary. That made me use only original compositions when searching for developmental traits in his notation practice in the three periods. The total distribution of entries in the four vocabulary groups also indicated a dramatic change entering the Serial period.

The change can also be seen in the distribution of Literary expressions in the three periods:

As well as in the distribution of Tempo vocabulary.

                                                 (Entries per work, no Tempo Character in the Serial period)

During the registration, I experienced that the work categories had very different numbers of entries. In all workgroups, except Voice ensemble, a dramatic reduction of vocabulary is registered in the Serial period. This new level can be seen in all four groups, articulation, dynamics, tempo, and especially Literary expressions.

In the Russian works, Stravinsky uses 189 different signs and expressions. There are 184 different signs in the Neoclassical works. However, there are only 95 signs and expressions in the Serial works (Including non-serial works). The Mean per work is nearly the same in Russian and Neoclassical works and less than half in the Serial works. The correlation coefficient between the vocabulary in Russian and Neoclassical periods and between Neoclassical and Serial periods are nearly the same, indicating that the basic notation is the same in all three periods.

The reduction in the Serial period is related to reduced/lack of adjustments, nuances, and other character expressions. As such, his Serial notation practice is much more in line with his ideas about objectivity and absolute music than his notation practice in the Neoclassical works, even though he promoted those ideas most intensively in the Neoclassical period.

Studying his scores as his way of communicating his musical ideas to the performers indicates a need for a critical comparison of his notation practice and his ideas about notation and music. Sometimes he seems to think of the notation as an object parallel to literary utterances, in line with Saussure's view of language as "a system of signs that express ideas". As part of his crusade for neoclassicism and objectivity, he embraced the idea of music notation as an a-historical practice. Add Stravinsky's lifelong occupation with rituals, and we have a background for his wish for music notation as a self-explanatory entity.

On the other hand, Stravinsky's constant concern about interpretation indicates an acceptance of the music notation as a system of signs that must be contextualised and interpreted. When Stravinsky was conducting, he could give far more expressive instructions than the notation indicated. (In the Studio session with Cathy Berberian, he pronounces the Russian words with an expressivity far more intense than in the notation.)

In the Serial period, Stravinsky reduces his vocabulary to such a degree that a performer might develop an interpretation that goes far beyond the notated score. Nevertheless, having Stravinsky's statements on performance/execution, most musicians would perform his Serial pieces as "werktreue" as possible. However, that might end up in playing "stupid notes," as Boulez called it, referring to performing single notes without understanding the composition's context.

Developing a craftmanship in serial composition, guided by Robert Craft, made old Stravinsky catch up with the development of musical styles in a way that came out as a reduced version of his communication to the musicians. Here he went even further searching for objectivity in the music than in earlier periods. Stravinsky's New Objectivity in the Serial period focuses more on the ideological dimension, the internal musical qualities reducing the vocabulary that refers to external characters like nuances, and he indicated tempo now only by Metronome numbers (the only objective indication in the music notation system.) "Composers combine notes. That is all." He said that after presenting a lot of visual impressions relating to the Symphony in Three Movements. The quote is more in line with his ideas of absolute music practised in his Serial period than for the symphony from the Neoclassical period. A consequence could be that when performing Stravinsky's works from the Neoclassical period, one should not use his literary utterances preaching objectivity as a criterium for reliability.

There is one problem with this, and that is Stravinsky's recordings of his music. They are often given a very dry soundscape towards an objet d'art, and he is very occupied with having correct/notated tempi. However, when commenting on three recordings of The Rite in Themes and Conclusions, he concludes that his recording is the best "Because the tempos are better on the whole and because there is more strength behind it". I would say that the tempos are indicated in the score but to get the strength behind it seems to be a result of the interpreter's/conductor's imaginary decisions mediated to the musicians. We can also find this part of his musicianship in the studio sessions, where he focused more on making music and allowed deviations from the score if he got the proper expression; that is, he focused on validity.

By loosening the connection between Stravinsky's ideological utterances and his musical works, we could reduce the incongruencies so often presented in the literature. So, even if my paper has given no clear answer, I hope my paper and the book could contribute to a discussion on whether Stravinsky had three periods but only two styles when performing his music: One covering the Russian and Neoclassical periods and one for the Serial works.


Dahl, Per: Modes of communication in Stravinsky's Works. Sign and Expression. (Routledge 2022)