 The primary phonological unit in many classical approaches is the phoneme. However, there is theoretical as well as empirical evidence that a unit smaller than the phoneme is more suitable. This unit is referred to as the distinctive feature. In order to understand today's system of distinctive features, this e-lecture introduces the history and the foundation of distinctive features. In a follow-up e-lecture about distinctive features, we will outline the advantages of a feature-based approach using today's feature system. So let's start with the history. There are three main approaches that have influenced the definition of distinctive features. The first is the theory of distinctive oppositions which goes back mainly to the Russian phonologist Nikolai Trubetskoy. The second is an acoustic approach which was first introduced by the Danish linguist Roman Jacobsen. And then there is the approach introduced by Noam Chomsky and Maurice Halle in their influential book The Sound Pattern of English. Let us look at these approaches in detail. The theory of distinctive oppositions goes back to Nikolai Trubetskoy in 1939. It was an attempt at classifying the properties of distinctive contrasts operative in natural languages. Trubetskoy defined a taxonomy of phonological contrasts that is of meaning-changing oppositions. There are oppositions within the system. There are relationships between opposite members. And there are types of oppositional extent. Thus Trubetskoy's approach was not a general theory of distinctive features. Nevertheless, he provided the framework for alternative views about phonological contrasts. So let us look at his taxonomy in detail. Oppositions within the system can be defined according to Trubetskoy in three ways. Segments stand in multilateral opposition if there is more than one phonetic property that keeps them apart. For example, in languages like Korean the system of plosives has always three members. Voiced unaspirated plosives, voiceless unaspirated plosives and voiceless aspirated plosives. There are two differentiating features, plus-minus voice and plus-minus aspiration. So a multilateral opposition. An opposition can be defined as proportional if the relation between its members is identical with the relation between the members of another opposition or several other oppositions within the system. The difference between T and D in English is identical with that of K and G in English. Thus we have a proportional relation. An opposition can be defined as isolated if no other segments stand in the same relation in the language. For example, the opposition between an alveolar approximate and an alveolar lateral consonant in English is isolated since no other segments stand in the same opposition. Opposite members can also be defined in terms of their relation or their opposition. In privateive oppositions one member carries a phonetic property which the other member lacks. In other words, privateive oppositions are relationships of the presence versus the absence of a feature, for example voicing. Oppositions in which the members are characterized by different degrees or gradations of the same properties are defined as gradual. This type of relationship holds for vowels where the property tongue height is a matter of degree rather than of binary. If the relationship between the members of an opposition can be considered logically equivalent it is defined as equipollant. A typical case is the opposition between English P versus T and T versus K. Unlike with vowels it is not possible to speak of a continuum from labial to vela and distinguish the consonants by means of degrees of backness. Rather, the different places of articulation constitute an equipollant relation. The extent of an opposition can also be defined in Troubet's Coise taxonomy. When two members of an opposition can occur in all positions the opposition is referred to as constant. This is almost the case as we will see in a second for English T and D. An opposition between segments is called neutralizable if there are contexts where only one of the opposition members can be found. A classic example of a neutralizable opposition comes from German where the opposition between voiced and voiceless consonants such as T and D disappears in final position. Or, and I will illustrate this right now, from present day English where the opposition between T and D is neutralized after S. Listen and see what's happening. Stick, stick, stick, stick, stick, stick, stick, stick, stick. And you see after S the opposition between T and D no longer exists. It is neutralized. Let's now turn our attention to the second approach towards distinctive features as defined by Roman Jacobson in the 1950s. Jacobson aimed at a description of the opposition's operative in languages. Together with his Swedish colleague Gunnar Funt and the American phonologist Maurice Halle he proposed a system of binary features now called distinctive that capture phonological contrasts rather than describe the phonetic segments themselves. But what exactly does binary mean? Well, a feature is defined as binary if it has two values. One of which is designated as plus the value and the other as minus the particular value. In many cases a binary approach is only phonologically significant in those oppositions which Troubet's Coy termed privative. For example, the feature nasal is binary phonologically. Phonetically, however, it is a matter of degree since some speech sounds can be more or less heavily nasalized. Probably due to the availability of the sonar graph, later referred to as spectrograph, which enabled phonologists to analyze speech sounds acoustically, Jacobson's features were acoustic in character. He defined these three feature groups, major class features, features for vowels, features for consonants and secondary articulation features. Let us look at Jacobson's acoustic features in detail. Now, the first group of features Jacobson defined are referred to as major class features. They have been introduced to distinguish the major classes of segments, vowels, consonants, and segments somewhat in between nasals, liquids and glides. However, there is disagreement about the number and type of major class features. Jacobson introduced the features consonantal and nasal. Later, as we will see, further features were added. So, in Jacobson's approach, all vowels were assigned the major class feature minus consonantal. To account for the parameters tongue height, tongue position, and lip rounding, additional features based on acoustic measurements came in. Here they are. They look complicated at first sight, but they aren't. The feature diffuse relates to all segments whose energy concentration can primarily be found in a narrow region central to the frequency spectrum. This defines vowels such as e and u, as well as most non-alveolar consonants as plus diffuse. The feature compact relates to all segments with low energy concentration in the narrow central regions of the spectrum. This defines the low vowels as compact. The feature grave relates to all segments whose energy concentration can primarily be found in a narrow region of the lower frequencies of the frequency spectrum. This defines all back vowels as plus grave. And the feature flat down here relates to all segments with weak or low energy concentration in the upper regions of the frequency spectrum. This defines most back vowels as plus flat. In Jacobson's feature system, all segments specified as plus consonantal qualify as a consonant. This feature distinguishes sounds with low energy and substantial obstruction from non-consonantal sounds. Among the features assigned to such segments are plus minus diffuse, plus minus grave, and plus minus flat. They were also used for vowels as we've just seen. Additional features were primarily articulatory in character. The feature voice correlates with vocal chord action. A speech sound is voiced and has the feature plus voice if the vocal chords vibrate. It has the feature minus voice if the glottis is open. Nasal sounds are produced by lowering the velum and allowing the air to pass outward through the nasal cavity. Non-nasal or oral sounds are produced with the velum raised to prevent the passage of air through the nose. Continuant sounds involve a primary constriction allowing the air to flow through the mid-sagittal region of the vocal tract. Sounds produced with a sustained occlusion are defined as minus-continued, hence vowels, glide trills and friccatives are plus-continued. Plosives and laterals are minus-continued. The feature plus-minus-continued thus by and large correlates with the distinction between plosives and non-plosives. The feature strident or sharpness of noise was introduced by Roman Jacobson to distinguish consonants involving high energy at high frequency from consonants involving low energy at low frequencies. Thus it distinguishes noisy consonants like ss from more mellow consonants like shh. So alveolar friccatives are classified as plus-strident whereas dental and labiodental ones and post-alveolar ones are defined as minus-strident. In order to capture aspects of secondary articulation such as labialization, palatalization, etc. The following acoustic features were introduced. Flat, a feature which refers to the weakening of upper frequencies, sharp, the strengthening of upper frequencies and checked the higher rate of energy discharge. The most influential approach of our time is the sound pattern of English approach which is essentially a book that was published in 1968 by Noam Chomsky and Morris Halle and they defined distinctive features in two ways as units that capture phonological contrast and patterns of languages and as means that describe the phonetic content of segments. The sound pattern of English system can be divided into three general groups of features. Major class features, primary placement features and additional features. So let's look at them. Like Jacobson, Chomsky and Halle used the feature Consonantal, here abbreviated as Cons, to distinguish consonants from vowels. They defined all those sounds as plus-consonantal that are produced with some sort of constriction in the vocal tract at least equal to that found in fricative consonants. Sounds that have the feature Syllabic are those which constitute peaks of syllables. Non-syllabic sounds are those which are in the margin of syllables. Assuming a non-linear format of phonological representation, especially the onset rhyme split of the syllable, this feature becomes redundant since it coincides with the peak of the syllable. So here are some examples. Let's take the word cattle. C, which is the onset, can be defined as plus-consonantal minus-syllabic. In cattle would be typically minus-consonantal plus-syllabic, it stands at the peak. And then we have T, which is another consonant, which is plus-consonantal minus-syllabic. And since the final L, the alveolar lateral constitutes a syllable in its own right, it can be defined as plus-consonantal plus-syllabic. Speech sounds, and that's the last major class feature, can be defined as sonorant if the pressure of air behind the constriction is approximately equal to the ambient air pressure. According to their amount of sonority, all speech sounds can be ranked, where one is the highest degree of sonority and five the lowest. The three highest, that is vowels, liquids and nasals, are defined as plus-sonorant. The two lowest, fricatives and plosives are minus-sonorant. A set of primary placement features supplements the major class features. Vowels can be defined using four such placement features. If they involve a high tongue position, they are referred to as plus-high. If the tongue is in a low position, the feature assigned to them is plus-low. And if the back of the tongue is involved, they are specified as plus-back. Finally, the distinctive feature round is used for vowels that involve lip rounding. It distinguishes the primary set of cardinal vowels from the secondary set. Here are some examples. E is defined as plus-high, minus-low, minus-back, minus-round. A cardinal four would be defined as minus-high, plus-low, minus-back, minus-round. And O would be defined as minus-high, minus-low, plus-back, and plus-round. We still need a fifth feature, but that will be shown in our second e-lecture about distinctive features. Consonants are defined by two features that define them on the basis of their place of constriction. That is, from labial to uvula. The feature, coronal, which you can see here, refers to the activity of the blade of the tongue. This involves all those sounds that are produced between the dental and the palatal alveolar region. The expression coronal is defined as pertaining to the blade of the tongue. A more transparent term would thus be bladele, yet this one doesn't exist in English. For this reason, the feature coronal relating to the crown of the tongue has been chosen instead. The feature anterior is associated with all those coronal consonants which are produced with the primary constriction located at or in front of the alveolar ridge. That is, with labial, dental, and alveolar consonants. Here are some examples. Per number one would be defined as minus anterior, minus coronal. It's outside the coronal region. Te is typically coronal and it is located in the anterior region. Ke, a palatal plosive, is minus anterior but still coronal. And ke, a vela plosive, would be minus anterior and minus coronal. Additional features in the sound pattern of English approach were taken over from Jacobson. For example, the features continent and strident. Whereas continent correlates with the distinction between plosives and nonplosives by and large, the feature strident distinguishes noisy fricatives from less noisy consonants. And then Chomsky and Halle added the feature delayed release, which was introduced to distinguish plain plosives from African consonants. The latter ones, the Africans, involve a delayed release of the oral closure due to the friction component preceding the plosives. Thus, Africans were defined as plus delayed release. Now, the three approaches defined in this e-lecture are by no means complete. However, they served as the input to the distinctive feature definitions that are used in modern phonology. Today, no serious phonologist defines phonological rules without reference to the theory of distinctive features. In a second e-lecture about distinctive features, we will thus first of all look at those features that are used in modern phonology and will then show how phonological rules can be defined by means of distinctive features. So, see you again in distinctive features too. Until then, thanks for your attention.