Academia.eduAcademia.edu

Figure 11 – uploaded by Giorgos Markopoulos

See full PDFdownloadDownload figure
the surface level. Finally, the candidate contains a sequence of adjacent vowels and thus violates *VV. In this case, the violation is represented by the integer number 1, as it does not

Figure 11 the surface level. Finally, the candidate contains a sequence of adjacent vowels and thus violates *VV. In this case, the violation is represented by the integer number 1, as it does not

Related Figures (85)

examples from Greek (in phonemic transcription): What we observe is that the masculine gender value is present in all four inflectional forms of the noun fixos ‘wall’, as opposed to number and case values, which vary and hinge on the
examples from Greek (in phonemic transcription): What we observe is that the masculine gender value is present in all four inflectional forms of the noun fixos ‘wall’, as opposed to number and case values, which vary and hinge on the
percolate to the first branching node, as illustrated in (8) (Lieber 1980: 53): In the Latin example above, although the features of the root am percolate to the non-
percolate to the first branching node, as illustrated in (8) (Lieber 1980: 53): In the Latin example above, although the features of the root am percolate to the non-
specification. The derivation that takes place in this case is exemplified in (13a—b): nerge of mP with the functional head of Number, while verbal structures result from the
specification. The derivation that takes place in this case is exemplified in (13a—b): nerge of mP with the functional head of Number, while verbal structures result from the
The question that arises at this point is how these features are “translated” into phonological information that can be processed at the PF level. In other words, how do we get from the tree structure in (14) to the surface realization tréxo? This brings us to the discussion of Vocabulary Insertion (V1), unfolded in the following section. ‘run.ACT.IPFV.NPST.1SG’ is the output of the following derivation process:* information that can be processed at the PF level. In other words, how do we get from the heads need to be assigned specific values. So, for instance, the Greek verbal form tréxo
The question that arises at this point is how these features are “translated” into phonological information that can be processed at the PF level. In other words, how do we get from the tree structure in (14) to the surface realization tréxo? This brings us to the discussion of Vocabulary Insertion (V1), unfolded in the following section. ‘run.ACT.IPFV.NPST.1SG’ is the output of the following derivation process:* information that can be processed at the PF level. In other words, how do we get from the heads need to be assigned specific values. So, for instance, the Greek verbal form tréxo
We are now in position to see how the structures and the operations we have assumed so far are represented in a nominal form in Greek. Given the tree structure in (15b), in order to lerive the word for “humans”, namely the nominative plural form of the noun that denotes a
We are now in position to see how the structures and the operations we have assumed so far are represented in a nominal form in Greek. Given the tree structure in (15b), in order to lerive the word for “humans”, namely the nominative plural form of the noun that denotes a
of the DP, as illustrated in the examples (18-20):
of the DP, as illustrated in the examples (18-20):
Table 1. Masculine nouns in -os/-as and neuter nouns in -o/-a cannot be predicted from either the gender value (Table 1) or the form of its ending (Table 2).
Table 1. Masculine nouns in -os/-as and neuter nouns in -o/-a cannot be predicted from either the gender value (Table 1) or the form of its ending (Table 2).
Table 2. Masculine nouns in -as and feminine nouns in -a
Table 2. Masculine nouns in -as and feminine nouns in -a
Table 3. Verbs with irregular aspect forms in the stem-final consonant (z — k). Interestingly, vowel alternations are not uniform across
Table 3. Verbs with irregular aspect forms in the stem-final consonant (z — k). Interestingly, vowel alternations are not uniform across
Table 4. Masculine nouns in -as/-is and feminine nouns in -a/-i
Table 4. Masculine nouns in -as/-is and feminine nouns in -a/-i
ubject, SUBJ) in a higher position within the same TP, as illustrated in the tree structure ir Furthermore, it also bears the [+inferior] value, because there is always another DP (i.e., the
ubject, SUBJ) in a higher position within the same TP, as illustrated in the tree structure ir Furthermore, it also bears the [+inferior] value, because there is always another DP (i.e., the
structure we assume for Greek nominals is the following: According to the discussion in the previous sections, the underlying morphosyntactic
structure we assume for Greek nominals is the following: According to the discussion in the previous sections, the underlying morphosyntactic
However, this is not the whole picture. After the syntactic output depicted in (58) is sent to PF, it undergoes certain modifications driven by specific well-formedness requirements of Greek morphosyntax. To explain, in Greek the n head that merges with Number triggers the insertion of an adjacent Theme node (TH), indicated by the dashed line in (59b) (see Embick & Noyer 2007 for a similar account for Latin nominal declension). Crucially, this process takes place at a post-syntactic level and has a bearing only on the PF interpretation. akes place at a post-syntactic level and has a bearing only on the PF interpretation.
However, this is not the whole picture. After the syntactic output depicted in (58) is sent to PF, it undergoes certain modifications driven by specific well-formedness requirements of Greek morphosyntax. To explain, in Greek the n head that merges with Number triggers the insertion of an adjacent Theme node (TH), indicated by the dashed line in (59b) (see Embick & Noyer 2007 for a similar account for Latin nominal declension). Crucially, this process takes place at a post-syntactic level and has a bearing only on the PF interpretation. akes place at a post-syntactic level and has a bearing only on the PF interpretation.
However, this is not the whole picture. After the syntactic output depicted in (58) is sent to to Number, yielding a complex head (an M/-Word), as illustrated in (58): PF, it undergoes certain modifications driven by specific well-formedness requirements of
However, this is not the whole picture. After the syntactic output depicted in (58) is sent to to Number, yielding a complex head (an M/-Word), as illustrated in (58): PF, it undergoes certain modifications driven by specific well-formedness requirements of
while its number and gender features are valued via copying from the Number and the n node
while its number and gender features are valued via copying from the Number and the n node
movement and post-syntactic feature insertion have taken place, is illustrated in (67): ‘ollows: the n[+human] head selects the root V#FISH and then merges with the Number head
movement and post-syntactic feature insertion have taken place, is illustrated in (67): ‘ollows: the n[+human] head selects the root V#FISH and then merges with the Number head
faithful one, namely psarades, (72a). In this case the relevant constraints are the following: rom the violation of DEP and MAX, violate also the markedness constraints *VV and *C.
faithful one, namely psarades, (72a). In this case the relevant constraints are the following: rom the violation of DEP and MAX, violate also the markedness constraints *VV and *C.
complete syntactic structure we put forward for the feminine noun xipsteru (nominative proposed in (66), we need to further address two issues: first, why -ao.6d0.s is absent from the
complete syntactic structure we put forward for the feminine noun xipsteru (nominative proposed in (66), we need to further address two issues: first, why -ao.6d0.s is absent from the
What the tableaux (84-85) reveal is that the phonological processing of an underlying representation may yield a significantly different output. Despite this mismatch, all the alternations the phonological input undergoes are driven from general principles and well- formedness requirements of the phonological component, and thus can be easily captured by a grammatical model that makes use of such principles and constraints. In other words, in What the tableaux (84-85) reveal is that the phonological processing of an underlying by doing so it avoids the violation of the *VV constraint: representation may yield a significantly different output. Despite this mismatch, all the
What the tableaux (84-85) reveal is that the phonological processing of an underlying representation may yield a significantly different output. Despite this mismatch, all the alternations the phonological input undergoes are driven from general principles and well- formedness requirements of the phonological component, and thus can be easily captured by a grammatical model that makes use of such principles and constraints. In other words, in What the tableaux (84-85) reveal is that the phonological processing of an underlying by doing so it avoids the violation of the *VV constraint: representation may yield a significantly different output. Despite this mismatch, all the
n{+human] head has no THEME node; hence /i6/ is not inserted into the structure. are adjacent to Number). As a result, in the morphosyntactic structure of mandviko, (88), the surface level, but mainly because there is suprasegmental information that also needs to be
n{+human] head has no THEME node; hence /i6/ is not inserted into the structure. are adjacent to Number). As a result, in the morphosyntactic structure of mandviko, (88), the surface level, but mainly because there is suprasegmental information that also needs to be
insertion of the THEME node, the structure in (95): plural, besides the number/case exponent -a,*’ the common part in all forms is the sequence - mat-. We therefore conclude that the phonological representation of the exponent includes,
insertion of the THEME node, the structure in (95): plural, besides the number/case exponent -a,*’ the common part in all forms is the sequence - mat-. We therefore conclude that the phonological representation of the exponent includes,
The output of the insertion of the above exponents into the structure is xtipi ~ mato. >xponents for the morphosyntactic structure in (95). The full exponent list is given in (97
The output of the insertion of the above exponents into the structure is xtipi ~ mato. >xponents for the morphosyntactic structure in (95). The full exponent list is given in (97
position; otherwise it violates *Cj-son, -sib]}o and thus reduces significantly the H value of the output (98a). Furthermore, in the plural, its realization prevents the emergence of a hiatu: environment (/aa/); therefore the output xtipimata (99a) is more harmonic compared to
position; otherwise it violates *Cj-son, -sib]}o and thus reduces significantly the H value of the output (98a). Furthermore, in the plural, its realization prevents the emergence of a hiatu: environment (/aa/); therefore the output xtipimata (99a) is more harmonic compared to
provided in (103): singular form trapézi, along with the phonological exponent of each terminal node, is
provided in (103): singular form trapézi, along with the phonological exponent of each terminal node, is
form karékla, along with the phonological exponents for each node, is as follows: that the THEME exponent /a/ is not realized. This is because the phonological exponent of
form karékla, along with the phonological exponents for each node, is as follows: that the THEME exponent /a/ is not realized. This is because the phonological exponent of
Given the underlying structure presented in (109), the phonological component has three
Given the underlying structure presented in (109), the phonological component has three
Our next move is to examine how the phonological component deals with cases where the floating stress with specific linearization properties. These properties may be of two kinds.
Our next move is to examine how the phonological component deals with cases where the floating stress with specific linearization properties. These properties may be of two kinds.
constitutes the phonologically optimal choice JUIN pnono1opiCdl COMpPUtaAtLON, UWE TediuZauOn OL WlEse WOCHNCES, WHICH NECds d U.F OL xtra activity, proves more cost-effective compared to the realization of the default yenultimate stress, which, as we saw earlier, requires an activation level equal to 0.6. This is Illustrated in the tableau in (115-116), where the outputs @dlasa (115b) and ayord (116b) *xhibit a higher H value compared to @aldsa (115c) and ayora (116c) respectively. As for the nput kareklao.s (117), the default penultimate stress in this case has no competition and thus input kareklao.s (117), the default penultimate stress in this case has no competition and thus
constitutes the phonologically optimal choice JUIN pnono1opiCdl COMpPUtaAtLON, UWE TediuZauOn OL WlEse WOCHNCES, WHICH NECds d U.F OL xtra activity, proves more cost-effective compared to the realization of the default yenultimate stress, which, as we saw earlier, requires an activation level equal to 0.6. This is Illustrated in the tableau in (115-116), where the outputs @dlasa (115b) and ayord (116b) *xhibit a higher H value compared to @aldsa (115c) and ayora (116c) respectively. As for the nput kareklao.s (117), the default penultimate stress in this case has no competition and thus input kareklao.s (117), the default penultimate stress in this case has no competition and thus
Similarly, with respect to tense and agreement, we also follow the literature in featural specification of T is based on the +/—past distinction, while for Agr we assume thi unary person features [1], [2] and [3] and the binary number feature [+plural].*”
Similarly, with respect to tense and agreement, we also follow the literature in featural specification of T is based on the +/—past distinction, while for Agr we assume thi unary person features [1], [2] and [3] and the binary number feature [+plural].*”
advocated against a separate Voice head and we subscribed to Embick’s (1998, 2004) view that voice features are inserted post-syntactically in order to signal a specific structural configuration. However, taking into account that, as we showed earlier, voice and aspect are
advocated against a separate Voice head and we subscribed to Embick’s (1998, 2004) view that voice features are inserted post-syntactically in order to signal a specific structural configuration. However, taking into account that, as we showed earlier, voice and aspect are
be identified as verbalizers, namely as categorizing heads that attribute verbal status to a roo and -ar. Some representative examples are given in (132):
be identified as verbalizers, namely as categorizing heads that attribute verbal status to a roo and -ar. Some representative examples are given in (132):
INT[F] in (148a) and MAX and *C.C in (148b). did for the root ayapVo.5-, we assume that the root-final segment has partial presence in th violation of DEP it triggers is less significant compared to the combined violation of DEP and
INT[F] in (148a) and MAX and *C.C in (148b). did for the root ayapVo.5-, we assume that the root-final segment has partial presence in th violation of DEP it triggers is less significant compared to the combined violation of DEP and
silencing, since the latter yields a cluster that is not allowed by *C.C (/xs/, 149b). presence of a zero exponent (/@/); for instance, the assumed underlying representation for the
silencing, since the latter yields a cluster that is not allowed by *C.C (/xs/, 149b). presence of a zero exponent (/@/); for instance, the assumed underlying representation for the
violates *VV. 1varmonic than dreap- (151a), because the latter not only is heavily penalized by DEP but also 6! See Revithiadou et al. (to appear); see also Spyropoulos et al. (2017a,b) for a similar hypothesis employing output resulting from the combination of the root dreo.6p- and the exponent /ao.7/ is dreo.6ao.7p
violates *VV. 1varmonic than dreap- (151a), because the latter not only is heavily penalized by DEP but also 6! See Revithiadou et al. (to appear); see also Spyropoulos et al. (2017a,b) for a similar hypothesis employing output resulting from the combination of the root dreo.6p- and the exponent /ao.7/ is dreo.6ao.7p
‘I drown’, kop-o ‘I am cut’).
‘I drown’, kop-o ‘I am cut’).
The exact computations taking place during the output evaluation are depicted in (159):
The exact computations taking place during the output evaluation are depicted in (159):
As already drawn form the examples we have reviewed so far in the present chapter,
As already drawn form the examples we have reviewed so far in the present chapter,
rather simple as it involves only strong elements with non-defective AL. phonological grammar evaluates those underlying structures. The processing of /yrafo/ is resulting structures being /yrafo/ and /ayapVo.640.20/ respectively. Let us now look at how the
rather simple as it involves only strong elements with non-defective AL. phonological grammar evaluates those underlying structures. The processing of /yrafo/ is resulting structures being /yrafo/ and /ayapVo.640.20/ respectively. Let us now look at how the
which case it is arbitrary in nature:
which case it is arbitrary in nature:
n{+human] heads and neuter for bare n heads. The emergence of masculine gender as defaul
n{+human] heads and neuter for bare n heads. The emergence of masculine gender as defaul
The major gain of our approach is that we end up having a systematic correspondence feminine noun kniga is the realization of an n[+feminine] head: between nominal endings and gender features, as summarized in (178).
The major gain of our approach is that we end up having a systematic correspondence feminine noun kniga is the realization of an n[+feminine] head: between nominal endings and gender features, as summarized in (178).
having to resort to inflection class features. The rules we assume are the following:
having to resort to inflection class features. The rules we assume are the following:
™ The underlying defectiveness of /ao.6/ explains why, contra to what would be expected according to sonority
™ The underlying defectiveness of /ao.6/ explains why, contra to what would be expected according to sonority
by doing so it avoids the occurrence of a consonantal segment in word-final position: the *VV constraint is not pertinent, the phonological grammar opts to realize /ao.6/, because center of attention shifts to systems that employ the same sex-based gender criteria found
by doing so it avoids the occurrence of a consonantal segment in word-final position: the *VV constraint is not pertinent, the phonological grammar opts to realize /ao.6/, because center of attention shifts to systems that employ the same sex-based gender criteria found
The epenthesis solution can explain not only the distribution of the endings -e and -@ in
The epenthesis solution can explain not only the distribution of the endings -e and -@ in
Under Lampitelli’s analysis, Spanish roots must be lexically specified as to whether they
Under Lampitelli’s analysis, Spanish roots must be lexically specified as to whether they
So, how do we end up with two distinct realizations (pan and xéfe), from which only one
So, how do we end up with two distinct realizations (pan and xéfe), from which only one
option for the phonological component. realization would yield a phonological word ending in a non-coronal consonant. As we have
option for the phonological component. realization would yield a phonological word ending in a non-coronal consonant. As we have
Similarly, in the environment of plural, where the Number node is realized by the exponent the tableau in (195): /s/, the pronunciation of /eo.«/ is always preferred because it prevents the occurrence of a word-final positions. The impact of the constraint on the output selection process is evident in
Similarly, in the environment of plural, where the Number node is realized by the exponent the tableau in (195): /s/, the pronunciation of /eo.«/ is always preferred because it prevents the occurrence of a word-final positions. The impact of the constraint on the output selection process is evident in
To illustrate how our analysis unfolds, let us look at the syntactic derivation and the According to the above, the complete list of the proposed functional exponents is
To illustrate how our analysis unfolds, let us look at the syntactic derivation and the According to the above, the complete list of the proposed functional exponents is
After the insertion of the vocabulary items and their linearization, the morphosyntacti
After the insertion of the vocabulary items and their linearization, the morphosyntacti
° For the ease of reading, the following morphosyntactic trees and GHG tableaux focus only on the realizatic of the segmental structure of the exponents in question and do not take into account stress computation. Al. Morphosyntactic structure
° For the ease of reading, the following morphosyntactic trees and GHG tableaux focus only on the realizatic of the segmental structure of the exponents in question and do not take into account stress computation. Al. Morphosyntactic structure
A1.2.3. N[-human] nouns (singular)
A1.2.3. N[-human] nouns (singular)
* The postulated exponent here is a [thigh] subsegmental feature, which needs to merge with a [+vocalic] merge either with /Vo./, yielding an underspecified high vowel (GeorV/+hignjome), or with /o/, forcing the latter
* The postulated exponent here is a [thigh] subsegmental feature, which needs to merge with a [+vocalic] merge either with /Vo./, yielding an underspecified high vowel (GeorV/+hignjome), or with /o/, forcing the latter
‘4 The representation /"o/ indicates that the exponent consists of the segment /o/, which is associated with the
‘4 The representation /"o/ indicates that the exponent consists of the segment /o/, which is associated with the
Underlying representation (phonological input): /afereo 686/
Underlying representation (phonological input): /afereo 686/
Related topics:
PhonologyDistributed Morphology
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved