Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
SPPA 4030 Speech Science DESCRIPTIVE QUANTITIES In Speech Science, we measure and describe all physical correlates of talkers talking (e.g., variations in air pressure that talkers create; or, the movements of body parts that cause those pressure variations) in terms of three types of base units: one unit to describe length or distance; one unit to describe time; and, one unit to describe mass. These three kinds of unit descriptors are the building blocks for our entire vocabulary for physical description. As a rule, when we describe the physical act and effects of speech production, we tend to follow what is known as the cgs (centimeter-gramsecond) system, a fore-runner of the more modern mks (meter-kilogram-second) system often referred to as the System International or SI system that has been the standard for most physical sciences since about 1971. The fact that speech people often persist in old-fashioned (cgs) ways, and do not follow the modern (mks) standard, sometimes causes a little confusion. But, we might defend using the cgs system to describe speech on the grounds that cgs base units tend to match the magnitudes of physical objects and events associated with speech better than mks units. For example, many of the things we move when we speak -- the class of things we call articulators, including the tongue, lips, jaws, vocal folds -- are neither large nor dense. At least some articulators certainly aren't on the order of "kilograms heavy." (Remember: 1 kilogram ~ 2.2 pounds.)1 Instead, they are more likely to weigh only fractions of a kilogram. For example, both vocal folds together might weigh about 2/1000 of a kilogram (= 2 grams). Each lip might weigh 25-50 grams or so, and the soft palate probably weighs about the same amount. The human tongue might weigh 150 grams. We have to talk about the human jaw, combined with the associated tissues tied to it, before we begin to approach an object that is a “kilogram heavy.” 1 It is common to “confuse” weight and mass, and possible to “confuse” volume and mass. Technically, neither association is correct, though it is obvious that the former conceit has been applied in a foregoing argument about the “masses” of articulators. For the time being, excuse the loose talk. Some will say that perhaps the best way to think about mass is to think in terms of stuff. Your mass – the amount of stuff that is you – is the same on Earth and the Moon, though you will certainly weigh less on the Moon. Thus, grams (abbreviated with a lower-case “g”) are a dimensionally appropriate way to talk about most articulatory masses. It is also true that the things we move when we speak typically do not move great distances. Movements spanning meters are out of the question, for most speakers and speaking situations. Movements spanning a few hundredths (e.g., the lower jaw, when speaking loudly), or even a few thousandths of a meter (e.g., the tongue surface during so-called speech diadochokinesis), are the rule. Consequently, centimeters (abbreviated with the two lower case letters “cm”) are a more suitable measure of articulatory distance or length. And, we don't move whatever we move, when we speak, either for long periods of time, or at great speeds. Significant events in speech "last" on the order of seconds, or more often, parts of seconds. In fact, a fair case could be made that milliseconds (thousandths of a second) and millimeters (tenths of a centimeter) are better measures of speech time and distance than seconds and centimeters, though even speech scientists aren't quite that radical. Certainly when we talk about the speeds of speech movements (expressed in terms of change in the position of an articulator, per change in time), seconds (abbreviated with a lower case “s”) seem to be about the best time unit to use, far better than the more familiar time units of minutes or hours that we might apply to our day-to-day lives. The fastest speech movements, for example, are strikingly slow when we think about them in miles per hour. For your own amusement, you might want to work out an answer to Practice Problem 7, figuring out how fast 50 centimeters-per-second is in miles per hour. Fifty cm/s is about as fast as we ever move any articulator when we speak, even when we try to talk as fast as a speeding locomotive. Thus, in the cgs system, which we encounter most often in the speech science literature, the standard base measures of distance = mass time centimeter (cm), = gram (g) , = second (s). These units contrast with the mks or System International units now followed in most of physics, where the standard base measure of SPPA 4030 Speech Science distance = mass time meter (m), = kilogram (kg), = second (s). We can refer to any combination of base measures (i.e., any quantity, expression, or concept involving at least two kinds of base units) as a derived unit. We use many derived units when we describe physical aspects of speech production. One of the more important of these derived quantities is pressure (as in the sound pressure wave). In words, we define pressure as force per unit area. In unit terms, a definition of pressure will thus involve some reference to area (= distance squared), and force, and it is obvious that we must define the latter before we can finally generate a derived unit expression for pressure. What follows are some derived units that are particularly relevant for quantifying events of interest during speech production. DISPLACEMENT: Displacement is perhaps the simplest derived unit. It is the base unit of length or distance, with the added dimension of direction. Displacement is considered a vector quantity because it has both a direction (e.g. up, down, west, east etc) and a magnitude (how much). In contrast, a scalar quantity has only a magnitude. For example, 2 cm is a base quantity of distance. It is a scalar quantity. A distance of 2 cm in the along an axis perpendicular to the earth’s surface is a displacement. AREA and VOLUME: Length or distance is measured along a single spatial dimension. But we know that the world around us has three spatial dimensions and often we need to capture these extra dimensions when describing an object. Measuring something with two spatial dimensions yields an area and measuring something in three spatial dimensions yields a volume. Area is expressed as distance2 and volume is expressed as distance3. In the cgs system, area takes the unit cm2 and volume takes the unit cm3. As an added annoyance, volume is often expressed in terms of liters or milliliters (which is equal to cm3). Both area and volume are frequently used for quantifying speech events. For example, during phonation, the vocal folds open and close causing an oscillating change in the glottal area. When we inhale and exhale for speech, it involves moving a volume of air into and out of the lungs. RATE MEASURES: Often we are interested in how quickly something happens. This is expressed as a change in some unit per unit change in time. When the unit of interest is displacement, the derived unit is displacement/time. We call this velocity. In the cgs system, the unit expression is cm/s. Like displacement, velocity is a vector quantity. The magnitude of velocity is commonly called speed and is familiar to anyone who drives a car. The speedometer tells us how fast we are going, but not our direction of travel. When the unit of interest is volume, we use the term flow rate (or volume velocity). As the name in parentheses implies, flow rate refers to the volume of a fluid (in speech, usually air) that moves past some place of observation per unit time. The unit expression is therefore cm3/s, though speech scientists often (and perversely?) describe airflow rates in liters/second (which will then equal cm3/s * 10-3). In the speech literature, we find at least two symbols used to represent flow rate: a capital letter "U", or a lower-case "i", though sometimes the reverse symbols are used (e.g., either a lower-case "u" or an upper-case "I"). We worry about flow rates in speech, for example, when we describe speaking habits of the hearing impaired who seem to be forever "running out of air" as they speak. The subjective impression is that such speakers spend a good bit more air per syllable (or word, or some other unit of speech) than normal speakers do. Put another way, flow rates for hearing impaired speakers often seem to be too high. Flow rate is analogous to current in an electrical system. This last fact is a handy thing to know since electrical circuits are often used to “model” acoustical systems (like the speech production system). Acceleration is the change in velocity per unit time. This may be represented as distance/time/time or simply distance/time2. In the cgs system, the unit expression is cm/s2. Like velocity, acceleration has a sign and a magnitude. You were probably first exposed to the concept of acceleration in high school physics. Recall that the Earth’s gravity causes objects to accelerate toward it at about 9.81 m/s2 or 981 cm/s2. However, we return to driving a car for a more concrete example of acceleration. Acceleration informs us about whether an object (i.e. a car) is speeding up (positive) or slowing down (negative). Rate measures such as velocity (and speed), flow rate and acceleration may be measured over different time scales. If the time scale is the SPPA 4030 Speech Science same as the duration of the event of interest, measurement would yield an average velocity/flow rate/acceleration. However, you can use a very small time scale which if small enough will yield something we call instantaneous velocity/flow rate/acceleration. For example, if you drive to Chicago, you look at the speedometer and get an estimate of your instantaneous speed, or you can wait until you get to Chicago and use the distance and overall driving time to calculate your average speed. FORCE: The unit force in the cgs system is the dyne. By definition, this is the force that will accelerate a 1-gram mass at 1 cm/s2. Thus, cgs quantities of force (as per Euler's expression of Newton's Second Law: F = ma) will have a 'unit expression' dyne = (g * cm)/s2, or some equivalent. Since the acceleration due to the Earth's gravity, at sea level, is about 981cm/s2, so that 1 "gram-force" is about 981dyne. In the “big” mks world, the unit force is the newton, by definition the force that will accelerate a 1kilogram mass at 1 m/s2. Since force is derived from acceleration, it too is a vector quantity with both a sign and a magnitude. PRESSURE: Pressure is generally defined as force per unit area, and will therefore have a cgs unit expression of dyne/cm2, or equivalently, g/(cm * s2). Atmospheric pressure (often abbreviated Patm) is about 1.013 x 106 dyne/cm2 at sea level. Interestingly, in the speech literature, air pressures are most commonly described in terms of cm of water (cm H20) or less often, mm of mercury (mm Hg). These ways of describing pressure derive from a method for measuring Patm that involves a column of water (or mercury) in a tube, whose height is read in centimeters (or millimeters). Patm in these terms is about 1033 cm H20 (or, 760 mm Hg). As you will learn later, the typical driving pressures for speech production, generated within the respiratory system, are about +10 cm H20 (with respect to Patm), or 1% higher than the ambient pressure exerted by the atmosphere. It is generally convenient to think of one cm H20 as equal to 1000 dyne/cm2 -close enough to merit a cigar. Among some speech scientists, the (modern) fashion is to describe air pressures in terms of a unit called the kilopascal (kPa). [A pascal, abbreviated Pa, is the pressure generated by one newton, acting over a one-square-meter area. As you might guess, a kilopascal is 1000 Pa.] Roughly speaking, Patm = 100 kPa, and thus, 1cm H20 is about 0.1 kPa. All these expressions for pressure can make things a little confusing. No matter what expression we choose, we worry a lot about pressures in speech production. For example, we know that the vocal folds will not vibrate if the difference between air pressures in the trachea (below the vocal folds) and pharynx (above the folds) is too small. Solving the "problem of phonation" (i.e., making the vocal folds vibrate) means that speakers must solve the problem of maintaining an adequate pressure “drop” or difference, above and below the vocal folds. Pressure is usually represented by a capital letter "P", and is analogous to voltage in an electrical circuit. WORK: Work is accomplished when a force is applied to an object and causes it to move. Therefore, work is defined as the product of force and distance. In the cgs system, the unit of work is called an erg (dynes*cm). The mks system uses the term joule. Raising your mandible 5 cm requires more work than raising your jaw 2 cm. POWER: Power represents the amount of work accomplished per unit time. This is a rate measure and the standard unit of watt uses the mks system (joules/s). Raising your jaw 2 cm in 1 second requires more power than raising your jaw 2 cm in 2 seconds. INTENSITY: Intensity is power per unit area and typically uses the unit watt/cm2. This is not strictly cgs since the watt is uses the mks system. Work and power are described mainly because they are building blocks for the derivation of intensity. We will not routinely use work and power. However, intensity is a quantity that has relevance for understanding the loudness of sounds. Note that pressure and intensity are both expressed as units (respectively force and power) per unit area RESISTANCE: Resistance refers to energy “losses” (e.g., due to friction and heat) that might arise when a fluid like air moves across a surface (e.g., through some tube). Analogous to an electrical circuit, and following the spirit of Ohm's Law (E = I*R -- i.e., voltage E equals the product of current I and resistance R), acoustic resistance, typically represented by an uppercase "R", is defined as a pressure difference (e.g., between two points along a tube) divided by flow rate (e.g., through the tube length defined by those two points), and will have a SPPA 4030 Speech Science 'unit expression' of (dyne * s)/cm5, or equivalently, g/(cm4 * s). Sometimes, a lowercase "r" is used instead of an upper-case one to indicate resistance. It is quite uncommon in the literature to see acoustic resistances represented explicitly in terms of either derived base-unit expression [(dyne * s)/cm5, or g/(cm4 * s)]. Rather, resistance is most often referred to in terms of “acoustic ohms,” or in terms of pressure (cm H2O) divided by flow rate (in liters/sec, of all things)! It is not an accident, in quantitative terms, that this poses no great problem relative to the strict definition since (about) 1000 dynes per cm2 corresponds to a pressure of 1 cm H2O, and 1000 cm3 correspond to a liter. Generally, we worry about acoustic resistance when we worry about the length, cross-sectional area, and shape of vocal tract constrictions. For (potential) speakers with cleft palates, for example, a major problem for the reconstructive surgeon is to provide the patient some means of creating a large enough velopharyngeal resistance so that every utterance doesn't seem to "come out of the nose." USEFUL CONSTANTS: Density of air (customarily represented by lower-case Greek letter 'rho', or the character ρ, is often given as 1.14 * 10-3 g/cm3 (i.e., mass per unit volume), for moist air at 37 degrees centigrade. Velocity of sound (customarily represented by a lower-case "c"), in moist air at 37 degrees centigrade is approximately 3.5 * 104 cm/s. A more common figure, for cooler air (e.g., ca. 0 degrees centigrade) is 3.31 * 104 cm/s. Of course, in the mks system, this is the same as 331 m/s. BRANCHES OF PHYSICS AND THEIR ROLE IN SPEECH SCIENCE The quantities outlined above are helpful in generating an organized account of our physical world. In this class, we are interested in understanding the physical basis of speech production. In a sense, this class may be considered a course in applied physics. Therefore, we should recognize that our descriptions and theories of speech communication borrow heavily from various branches of physics. The branch of physics that deals with sound is called acoustics. Aerodynamics is the branch of physics that deals with air volumes, air flows and air pressures. The description of bodily movement (for example, the displacement, velocity and acceleration of our articulators) is a branch of physics termed kinematics. We learned in SPPA 205 that movement (kinematics) arises from the generation of forces within the body and from external forces such as gravity. The branch of physics that deals with understanding the forces that underlie motion is dynamics.