Part I: Background: Linear Prosody
Dimensions of Inequality Among Syllables
In English, prosody proceeds from the axiom that syllables are not all created equal; much of prosody derives from the time-plot of these inequalities. Stress is the most familiar dimension of such inequalities; we will quickly review others.
Pitch Degrees
Usually pitch is considered as the intonation curve. However, in many American communities, most notably in the black community, poems are recited with pitches articulated in tight-knit time patterns similar to traditional stress patterns. This is a very rich and different prosody that deserves to be studied. While pitch is purely acoustical, stress is a linguistic property quite difficult to define acoustically. Thus a pitch degree prosody is much closer to (literal) music, and is freer to use absolute musical time -- vs. the "linguistic time" of stress-degree prosody. Because not all phonemes carry pitch, a pitch degree prosody changes the phoneme sound structure balance. Where both pitch and stress degree prosodies occur simultaneously, incredibly subtle effects are possible.
Vowel Position Degrees
Robert Duncan (Duncan, 1973) explained "Tone Leading Vowels" (applied to Pound) as meaning: (1) The leading pure vowel of a diphthong plays a special role in sound reinforcement; (2) A sound is reinforced when you hear it again, but can also be reinforced when you don't hear it again. This second point leads to the idea that vowels cluster according to the position of the mouth; the time pattern delineating which cluster is active forms a prosody analogous to stress or pitch degrees.
Stress Degrees: The Classical Stereotype
The most familiar basis for metrics is the time pattern of stress degrees. Stress has been extensively studied in linguistics (e.g. Chomsky and Halle, 1968). The classical stereotype stress-degree prosody starts with an a priori inventory of templates of stress degree patterns (e.g. iamb, trochee, anapest, etc.). Scanning consists of matching templates to the poem; where repeated instances of the same template match, the line is said to "scan". `Foot' is profoundly ambiguous, meaning: (1) one of the templates. Here `foot' is an abstract concept which exists in advance of any particular poem. (2) The actual syllables in a poem matched by a template. Here `foot' is part of a living breathing poem, constituting a unit of rhythm intermediate between the syllable and the metric line. Poetics since the fifties and sixties have typically focused away from the a priori (Olson, 1967), (Ginsberg, 1971) making many poets profoundly uncomfortable with a template approach, resulting in a neglect of the concept of an intermediate unit of metrics which is not a priori and is part of the poem itself. We now give a new concept of an intermediate unit of meter, avoiding the a priori. To avoid confusion, we will abandon the word `foot', and instead use the term `measure'.
Bonding Strength
Another inequality dimension is bonding strength: the degree of attraction of a syllable to the syllables on either side -- i.e. the naturalness of an artificially injected pause compared to the the poet's "normal" recitation. Bonding strengths will differ; by collecting syllables with high bonding strength one obtains a measure. It must be emphasized that measure boundaries are relative to a particular recitation -- presumably the poet's. A printed text may not give sufficient information. Here, scanning consists of empirically identifying measure boundaries, rhythmic line boundaries, and then attempting to discern any possible regularities in measure construction. (As opposed to considering a poem as "written in" a meter -- chosen consciously and a priori).
The "Standard Measure"
Prose can also be scanned. English prose measure boundaries tend to be constructed as follows: (1) a measure has exactly one major stress, which (2) tends to end the measure, but: (3) any following unstressed syllables out to the end of a major grammatical unit will be included in the measure. Such measures may be called standard measures. Not all poetic measure boundaries will be standard: Robert Creeley is well known for having non-standard measure boundaries. When Creeley's poems are actually scanned, many line-ending measure boundaries are non-standard, but the internal measure boundary in two-measure lines is standard -- the celebrated Creeley line-break really is a line-break not a measure break. The non-standard measure boundaries are very prominent, but the internal standard measure boundaries are much more subtle. This line structure may be described as an offset structure: line endings clearly articulated by sound, but grammatical structure proceeding from mid-line to mid-line. This structure goes back in prosody at least to Anglo Saxon times.
Part II: Non-linear Prosody
Bonding Strength is Spatial
Bonding strength is the attraction of a syllable to adjacent syllables. "Adjacent" is a spatial concept; thus bonding strength may be interpreted spatially -- in any topology, including a non-linear one. Bonding strength could be described as the tendency to resist artificial injection of space rather than a pause (time).
A Review of Hypertext Structure Terminology
We first review briefly the hypertext activity framework introduced in (Rosenberg, 1996, "Structure"). By hypertext we mean a text containing embedded interactive structure operations: interactive devices which trigger activities, such as the hypertext link. Many other types of activity are possible. This author's work uses simultaneities: layers of words atop one another, individually revealed by moving the mouse. Research hypertexts have used sets, relations, piles and lists. The embedded activities are e.g. following a link, opening or closing a pile or simultaneity, etc. Such activities are called actemes. Following a link may be described as a disjunctive acteme (in logic, disjunction is an or): one may choose among several alternative links. Other actemes are conjunctive: a simultaneity consists of all of its layers.[1]
Hypertext text typically appears in units called lexia (Landow, 1992) -- e.g. the unit of text at either end of a link. Often (though not inevitably!) the lexia has a linear internal structure. (The concept of lexia is extremely problematical; see below.)
During hypertext navigation, activities will (hopefully) cohere into units called episodes. For node-link hypertext, the episode will probably be all or part of a path. Some activities won't resolve into any episode: an activity might be unintentional, the user may backtrack, etc. Thus, episode is not simply history. Sometimes the user may not "have" an episode at all, but instead is foraging for an episode. Episode is an emergent concept; it emerges retroactively -- ideally through use of a gathering interface. Gathering interfaces are still in a primitive state; most simply gather "bookmarks" rather than assembling a full picture of hypertext activity.
Prosody Within the Lexia
The lexia may be (alas!) structured linearly; thus within-lexia prosody embeds traditional linear prosody. Not much need be said here. However, the lexia need not inevitably be linear. (See (Moulthrop, 1992), (Rosenberg, 1994).) We now consider within-lexia prosody for a non-linear lexia.
Figure 1 shows a screen from (Rosenberg, 1996, Diffractions), which can be read in multiple ways: (1) Polylinearly: words in the same font can be read as a linear skein, beginning with the capitalized word. (2) Graphically clustered phrase fragments can be "read in snatches", wandering about the screen associating groups of words in whatever way seems to work. As discussed in (Rosenberg, 1994), even (1) poses difficult questions for the concept of lexia: is "the lexia" the entire screen, or one of the skeins? A lexia might be described as whatever is visible absent input to the computer. In this case the entire screen should count as one lexia. What happens as the eye moves from one phrase to another? Is this time which "doesn't count" -- a time out in which there is no prosody? If so, time units within the skeins may be described as disengaged from one another. Or perhaps the prosody of a skein helps determine when the next phrase begins, making the time between skeins definitely part of the prosody.[2] Between-phrase prosody is inherently ambiguous in a polylinear lexia. Of course when the poet "recites" such a lexia, a choice will in fact be made, posing a difficult dilemma.
Figure 1
These issues are even more difficult in case (2) above. What is the prosodic relationship of clusters read by "visual wandering"? Here the reader may have an impression the words disengage from time altogether, with prosody becoming entirely spatial.
Prosody Through the Episode
Prosody need not be confined within the lexia. First, "text" may occur in the devices of the hypertext mechanism itself, such as a menu of outgoing links. What role does such a menu play in prosody? It might be considered as a text object in its own right. In (Larsen, 1996) poems are constructed from assembled link names. But here the link name menu simply constitutes a different form of lexia. Another approach is to consider a link name as a "prosody channel" connecting the text at either end of the link. In hypertext poetry, prosody is just as valid a means of choosing a link as semantic or logical criteria. Bonding strength can operate through the link, and may even be the basis for choosing a link in the first place. The relationship of prosody within the lexia to prosody through the link may be described as "two-dimensional" prosody. (Or even three-dimensional, if the lexia is spatial.)
In conservative hypertext, "the document" is the lexia; the interactive devices are merely navigational aids. A more radical approach treats the episode as a virtual document (Rosenberg, 1996, "Structure"): the "center of gravity" is no longer within the lexia, but in what emerges through activities. Ultimately, meaning -- and even syntax (Rosenberg, 1996, "Structure") -- become a function of the episode. What are the prosody implications of the episode as a virtual document? If we consider the episode as structured linearly by time, then prosody within the episode seems little different from linear prosody. But the user has chosen the interactions. In the disjunctive case the user has chosen an alternative among available actemes, in the conjunctive case the user has chosen the order of operating actemes. The user has complete control over how much time is spent in any given place in the hypertext and how much repetition occurs. This gives the prosody of the episode a high degree of indeterminacy compared to the prosody of the lexia.
Must the episode be linear? (Rosenberg, 1996, "Structure") argued that the episode is structured by the results of using a gathering interface. A gathering interface is a hypertext the user constructs of gatherings from the hypertext being read. It may use alternative spatial or conjunctive structuring methods, yielding a prosody of limitless complexity.
How Does Time Run in a Non-linear Poem?
Multiple concepts of time operate at once. (1) Usage Time is like a "raw tape" of what the user does. Even in the linear case this can be tricky. Consider isochrony, the tendency of stressed syllables to form a regular musical beat. Even when the acoustical correlates of stressed syllables are not evenly spaced, the stresses may so heavily influence our perception of time that they become our measure of time. Do interactive devices "become" the measure of time in an interactive poem? As hypertext is extended further into the fine structure of language, this may happen. Does usage time include all the "accidents" -- backtracking, overshooting a scroll-bar, etc?
(2) Gather Time is time spent constructing and reading the results in a gathering interface. Gather time may start and stop -- while foraging for episode gather time may be described as having stopped. Given a spatial gathering interface, is gather time "running" as one changes spatial relationships of gathered elements? Some type of time is running. How does the relationship of the reader to gathered phrases on a screen map to syllables during phrase manipulation? Is the time spent moving a phrase "mapped to" all the syllables at once? Can usage time work in this same way, given the right interface? Suppose words are laid out using graphical methods so that the eye associates all of the words together as a single object "all at once". Is time spent contemplating this object "suspended time"? Is it "autonomous time": the word object having its own concept of time, not necessarily reconcilable with the time concept of other objects present? Perhaps it resembles a loop: the words, once examined, "keep on playing".
A conjunctive structure consists of all of its components resolved into a single whole. What is the time relationship among these components? Usage time for each component may be described as equivalenced with that of the other components. In this author's simultaneities, usage history will resolve an order (under user control) in which the simultaneity elements were encountered. They aren't literally simultaneous, as in simultaneous voices, but the term `simultaneity' is meant to convey the idea that these units of time should be treated as equivalent. Equivalenced time is a natural correlate of the concept of autonomous word objects endowed with behavior. For another view of equivalenced time, see (Scalapino, 1997): "differentiation paired continually so as to occupy the same time".
Conversely, time units may be completely disengaged, their mutual time relationship being completely null. Juxtaposition is the null structure, or "structural zero" (Rosenberg, 1993); it is the crux of abstraction, clearly important in all of the arts for decades. In a hypertext, separate episodes may be time-disengaged even though usage time for one episode precedes usage time of another. Consider two memories, of incidents whose relative time one cannot determine. It doesn't matter in which order the memories were recalled; their time relationship is unresolved.
In a hypertext, time itself may become spatialized. An interactive multimedia device may permit playing a sound or movie with "its own" timeline which may actually show on the screen as a control -- anchored at a particular location. The hypertext as a whole will have no such timeline. Even for text, with no formal "player object", the entire text object may be anchored at a specific location. Travel through linear text is accomplished by reading linearly -- notwithstanding other ways of navigating a printed text, such as tables of contents, indices, footnotes, etc. In a hypertext, the specific interactive devices are likely to be used for travel, leaving the lexia as an anchored spot which "doesn't move". Thus the lexia, even if linear, is an anchored linearity and so is its time.
Multiuser Time
Thus far we have taken a "single-user" perspective, with a single reader of a work and a single (even if collective) author. Multiuser computer games are common; surely we will see multiuser literary works. (See Cayley, 1997, section [21].) Units of multiuser time are not necessarily resolvable from one user to another. Prosody events are typically passages over particular points in a poem -- syllables, line breaks, etc. With multiple concurrent readers in the same space, relative synchronization of these events may not be possible. Disengaged time is not metaphorical here, but a literal description.
The brain is a massively parallel system. A simple act of seeing involves substantial processing by each retina, even before the signals reach the brain. Might the "single-user" model not be correct even for a single reader? Is the brain itself perhaps "multiuser"? This is the question posed by (Dennett, 1991). Dennett devised a theory of consciousness as a parallel "gang of demons". Perhaps there are centers in the brain which act as "time disengaged actors" even for a single mind. Hypertexts can render this concept external and tangible. The questions raised for prosody have only begun to be asked. Much of prosody concerns reinforcement of sound events by earlier related (or differing! -- note Robert Duncan's comment above) sound events. How does reinforcement work with no way to determine which event occurs first? How does reinforcement operate across disengaged units of time? What happens to these time disengagements when the poet "recites" -- and how in general does a poet perform such a work?
References
Cayley, John. "HYPERTEXT/CYBERTEXT/POETEXT". This volume.
Chomsky, Noam, and Halle, Morris. The Sound Pattern of English. New York: Harper and Row, 1968.
Dennett, Daniel C.. Consciousness Explained. Boston: Little Brown and Company, 1991.
Duncan, Robert. Personal conversation. Circa 1973.
Ginsberg, Allen. Improvised Poetics. San Francisco: Anonym, 1971.
Landow, G. P.. Hypertext: The Convergence of Contemporary Critical Theory and Technology. Baltimore: Johns Hopkins University Press, 1992.
Larsen, Deena. Samplers. Computer Software. Watertown, MA: Eastgate Systems, 1996.
Mac Low, Jackson. 22 Light Poems. Los Angeles: Black Sparrow Press, 1968.
Moulthrop, Stuart. "Shadow of the Informand: A Rhetorical Experiment in Hypertext". Perforations 3. Atlanta, GA: Public Domain, 1992.
Olson, Charles. "Projective Verse". Human Universe and Other Essays. New York: Grove Press, 1967.
Rosenberg, Jim. "Openings: The Connection Direct". http://www.well.com/user/jer/openings.html. Liner notes included in Intergrams. Computer Software. Watertown MA: Eastgate Systems, 1993.
Rosenberg, Jim. "Navigating Nowhere / Hypertext Infrawhere". SIGLINK Newsletter 3, 3, December 1994. http://www.well.com/user/jer/NNHI.html.
Rosenberg, Jim. "The Structure of Hypertext Activity". Hypertext `96. New York: ACM, 1996.
Rosenberg, Jim. Diffractions through: Thirst weep ransack (frailty) veer tide elegy. Computer Software. Watertown MA: Eastgate Systems, 1996.
Scalapino, Leslie. "The Radical Nature of Experience". This volume.