Prosody refers to all other extra-linguistic information or anything that isn't a word. In layman's terms, prosody is about how you say something, not what is said. When it comes to spellcasting, prosody is the aspect that covers stress/emphasis and intonation.

Stress is a good place to start as it involves several extralinguistic features at once, these being pitch, volume and duration. Stress is a part of Metrical Theory which explores the importance of rhythm and meter and the way it affects how we speak.To understand stress, we must first understand syllabic structure.

Syllablic Structure

Syllables are tricky things as they can be difficult to define from one language to another. Essentially, they refer to a resonant peak or the point where the sound is the clearest and strongest. This peak is usually on the vowel, where the sound is least constricted by the vocal tract. This is referred to as the nucleus of the syllable. Anything before the nucleus is called the onset, and everything after is called the coda. Refer to the diagram below for reference.

Some languages prefer open syllables where there is no coda following the nucleus. An excellent example of this is Japanese. This predictable structure lends itself to a rhythmic or melodic quality that is very pleasing for incantations. There are a few potential theories as to why this is.

The first is Magical Meter Theory. This assumes that magical energy (like most physical energies) is composed of waves that pulse in a distinct rhythm. The closer the incantation aligns with this pattern, the more effective or powerful the incantation will be. This theory is well received among languages that use stress and meter, but does not account for other languages quite as well.

The second theory addresses the same phenomena with the proposed existence of moras. Moraic Weight Theory assumes that magic is composed of microscopic particles called moras. These moras tend to converge in groups which creates micro-pockets of magical energy. Through the use of incantations, syllables with varying sonic 'weight' can align with the magical energy much like a lock and key. This theory better accounts for languages without stress and meter but is still heavily contested as little is yet known or understood about moras and their groupings happen seemingly at random.

Since both theories have obvious drawbacks, it can be difficult to predict with any level of consistency which incantations will work. In order to prevent excessive danger caused from trial and error, magilinguists use optimality theory to compare the various linguistic forms of potential charms to select the candidate most likely to succeed.

For example we have 'Episkey' that has three vowel sounds ie. sonorant peaks/syllables.

The syllables can be divided in a number of different arrangements seen below.

All three lack an onset in the first vowel, while only the two close their codas. This leaves the first option as the winner, even though its final syllable is heavier than the others.

Meter

With this in mind we can return to stress and meter. In order to produce a rhythm, there needs to be a beat and off-beat. This is created by pairing syllables into units called feet. When the beat appears first in the pair, this is called trochaic meter ie. (stress-unstress). When the beat is on the second, the meter is iambic ie. (unstress-stress)

Let's use a simple spell as an example.

('lu.mɔs)

Lumos

As far as notation is concerned, the syllables are separated by periods and two syllables together are enclosed in brackets. The stressed syllable is marked by an apostrophe at the beginning of the syllable.

We see this particular spell has two syllables, the first being stressed, therefore this spell has trochaic meter. Easy enough. Let's try something a bit more difficult.

*(ɛks.'pəl).(i.,ɑʳ).mʌs ɛks.('pəl.i.)(,ɑʳ.mʌs)

Expelliarmus Expelliarmus

In this example I have provided two potential interpretations of the same outcome. The stress is placed on the second syllable of a five-syllable word.

When the syllables are paired from the beginning of the word, the stress falls on the second syllable making the meter iambic. However, If this is reversed and the syllables are paired from the end, you will see that the meter is actually trochaic. Given what we know of the language from the previous example, we can assume then that the second interpretation is correct.

This can even extend across words and into phrases.

(wɪn.'gɑʳ).(di.ʌm) (lɛ.vi).(,ɔu.sɑ)

Wingardium Leviosa

The stress appears on the second syllable of the first four-syllable word, but it appears on the third syllable of the second four-syllable word. Why is this the case?

Latin has a fixed penultimate stress meaning the emphasis is always on the second to last syllable. This is confirmed in 'Leviosa' where the stress is placed before the final syllable. 'Wingardium' is slightly more complicated, as you would expect the stress to fall on [di], however the previous syllable presents more syllabic weight therefore the moras would draw the stress to the heavier syllable. Another possible explanation is that the dipthong [iʌ] behaves as a single vowel and therefore would allow the stress to rest on the previous syllable.

Intonation

In combination with rhythm is also pitch. Pitch is generally calculated by frequency in Hertz. In English, we see pitch and intonation mostly as a function of sentence level prosody ie. asking a question versus making a statement. However, even stress includes a natural increase in pitch and volume making it a very important linguistic marker. In Latin spellcasting, intonation generally goes from high to low over the course of the spell but there is little evidence to prove that it has much of an effect either way.

Pitch is much more important to Asiatic and tonal languages that use pitch to distinguish minimal lexical pairs. There is a famous example in Chinese where [ma] which can mean five different words depending on which of the five tones is used.

When using lexical tone, there are two major systems that are generally used to serve this type of function:

Dynamic tone - These types of tones move from one pitch to another. They typically have fewer levels (high/low) as the relative movement is the most important feature of the sound. Some examples are rising, falling, static, rising-falling, falling-rising, etc.

Static tone - These tones have fixed levels ie. low/mid/high, and words or syllables are typically assigned a single tone at a time.

Perception

This is best explained through the relationship between categorical and continuous perception. Pitch for example can be perceived linearly like the scale on a violin. This is true of magic as well since 'will' can make a spell stronger or weaker.

Categorical perception is more like a piano where a specific key is assigned a specific sound. It is important to recognize these two methods as they work together. An incantation may have fixed sounds but there can also be a certain level of allowed variance. Allow me to explain.

Acoustic Theory

Sound is a kinetic disturbance in the air, meaning air is being pushed and pulled as sound is carried from one place to another. The ear has a mechanism inside that convert those subtle changes in air pressure into electrical signals in the brain. This is where the sounds are mapped based on pitch, volume, and quality. We will discuss this in further detail later, but for right now the most important thing to know is that hearing is as much in the brain as it is in the ear.

When we speak, air is pushed from our lungs through a cartilage ring in our trachea containing thin membrane-like folds. These are called the vocal folds, sometimes referred to as vocal chords. Depending on how much air we push through these folds will determine how much volume the sound has, and how close/tight the folds are will determine how high or low the pitch is.

As the air continues to travel forward through the larynx or voice box, it comes to the mouth where the jaw, tongue and lips form different shapes defining the quality of the sound. This is why we can tell the difference between an [a] and an [o] for example. Finally the resonant waves move through the rest of the face and nasal cavity, creating a vocal fingerprint if you will of the speaker. This is similar as to why violins and trumpets don't sound the same.

This brings us to a very interesting concept called the Speaker Normalisation Problem which wrestles with the fact that two speakers can produce very different acoustic signatures but still be understood as producing the same sound.

One of the ways, magilinguists account for these variances by studying formants. Formants essentially map how the resonant peaks of the word change over time. Much like the five-spellcasting elements, formants represent the most critical frequencies of a sound and how they relate to each other. Obviously, it gets much more complicated but the important thing to know is that formants are identified relative to each other and the successful alignment of the sounds is what allows incantations to align to the magical energy of the intended spell.

It is also important to recognise that formants are most visible on vowels which are the most resonant features of words. This is why many magilinguists believe there is a higher level of variance allowed in vowel pronunciation when performing an incantation or spell.

For reference, refer to the Vowel Quadrilateral, a chart mapping the allowable pronunciation quadrants for the various vowels.