Fortition
Fortition is a phonological process in linguistics whereby a speech sound, typically a consonant, undergoes strengthening, becoming more articulatorily or perceptually robust, often in prominent positions such as syllable onsets or word-initial contexts to enhance clarity and salience.[1] This process contrasts with lenition, its inverse, which involves weakening or increased sonority of segments; fortition is characterized by a decrease in sonority, shifting sounds toward less sonorous categories like stops or affricates from fricatives or approximants.[2] Common manifestations include gemination (lengthening consonants), spirantization reversal (fricatives to stops), or glide strengthening (e.g., /w/ to /β/ or /j/ to /dz/), frequently observed in diachronic changes across languages.[1] In theoretical frameworks like Optimality Theory (OT), fortition is modeled through interactions of markedness constraints that prioritize strong onsets and perceptual recovery, such as avoiding glottal stops in initial positions while permitting them in codas.[3] For instance, in Chamicuro, glottal contrasts are neutralized in onsets to favor non-glottal realizations, maintaining distinctions elsewhere, while in Awa Pit, word-initial onsets avoid glottal features to boost salience.[3] Diachronically, fortition appears in Indo-European languages, for example in the creation of geminates or strengthening in certain branches.[2] Fortition's typology aligns with obstruent hierarchies, where higher-sonority sounds (e.g., fricatives) are more prone to fortition into lower-sonority ones (e.g., stops), influenced by implicational scales based on articulatory and acoustic properties.[2] Examples abound in Romance languages, including Spanish word-initial stop realizations of /b, d, g/, and Greek dialects, where voiced geminates like /dd/ arise in Doric varieties (e.g., Cretan δικάδδοι).[1] These processes underscore fortition's role in balancing fluency and perceptual distinctiveness, contributing to sound pattern universals across synchronic and historical linguistics.[3]Definition and Basics
Definition
Fortition is a phonological process characterized by the strengthening of a consonant, whereby it increases in articulatory stricture and moves toward a less sonorous form on the strength hierarchy, such as transitioning from an approximant to a fricative or from a fricative to a stop.[4] This change enhances the consonant's perceptual salience and articulatory effort, often occurring in prominent positions like syllable onsets.[3] Key characteristics of fortition include the enhancement of features such as voicing, aspiration, or degree of closure, which collectively make the segment more "fortis" or robust compared to its weaker variants.[4] Unlike weakening processes like lenition, fortition reinforces consonantal properties to avoid marked or perceptually weak forms, such as glottal stops in onset positions.[3] The term "fortition" derives from the Latin fortis, meaning "strong," and has been employed in phonological literature since the early 20th century to describe these strengthening phenomena.[4] In basic phonological notation, it is represented as changes like /w/ → [β] (approximant to fricative) or instances where fricatives develop into stops, such as underlying /s/ → in certain historical contexts.[4]Relation to Lenition
Fortition serves as the phonological inverse of lenition, a process characterized by a reduction in the degree of oral constriction, such as the change from a stop to a fricative.[5] In contrast, fortition involves an increase in constriction, strengthening consonants from weaker to stronger articulatory realizations, thereby counterbalancing lenition's weakening effects within phonological systems.[5] This oppositional relationship underscores fortition's role in maintaining consonantal contrasts that lenition might otherwise erode. The functional opposition between fortition and lenition is prominently conditioned by prosodic position: fortition typically strengthens consonants in prominent contexts, such as word-initial or onset positions of stressed syllables, while lenition weakens them in less salient environments, like intervocalic or coda sites.[6] This positional asymmetry reflects underlying principles of articulatory ease and perceptual prominence, where strong positions favor enhanced articulation to ensure clear signal transmission, and weak positions permit reduction without significant communicative loss.[6] The fortis-lenis distinction, often invoked to describe this gradient of consonantal strength, further highlights their complementary dynamics in phonological structure.[7] In evolutionary patterns of language change, both fortition and lenition emerge as natural tendencies driven by articulatory and perceptual pressures, with fortition frequently countering lenition's progressive weakening over time.[8] A canonical lenition chain progresses from stops to fricatives, then to approximants, and ultimately to deletion (e.g., /t/ > /ð/ > /ɹ/ > Ø), reflecting incremental reductions in stricture and effort.[7] Fortition reverses these stages, as seen in historical chain shifts where weakened segments in strong positions are reinforced, such as fricatives realigning to stops or approximants to fricatives, thereby restoring systemic balance and preventing total loss of contrasts.[6] Such reversals in chain shifts illustrate fortition's compensatory function in long-term phonological evolution, ensuring the perpetuation of consonantal inventories across generations.[6]Phonetic Mechanisms
Articulatory Changes
Fortition often manifests articulatorily through an increase in oral closure, where consonants transition from partial obstructions, such as fricatives, to complete ones, like stops, by applying greater pressure from the articulators against the vocal tract roof. This strengthening elevates the degree of stricture, making the consonant more obstructive to airflow. For instance, in Garifuna, the fricative /ʃ/ fortites to the affricate [tʃ], reflecting heightened closure at the alveolar ridge.[5] Voicing and aspiration shifts in fortition typically involve enhanced glottal tension, which can lead to devoicing or the addition of aspiration, rendering consonants more fortis. In Germanic languages, this process enhances unmarked stops by introducing a [spread glottis] feature, promoting voiceless aspiration as a marker of strength. Such shifts increase laryngeal effort, distinguishing fortis series from lenis counterparts through greater glottal abduction. Place assimilation contributes to fortition by strengthening consonants via anticipatory articulatory gestures that align with neighboring sounds, such as labialization before rounded vowels. In languages like Apinayé, the palatal glide /j/ fortites to [ʑ] before high front vowels, where the anticipatory tongue raise and lip protrusion intensify the constriction. This coarticulatory effect bolsters the segment's articulatory prominence without altering its core identity.[5] Within the articulatory phonology model, fortition is characterized by constriction gestures that become more overlapped and intense, amplifying their magnitude and temporal coordination. Seminal work describes these gestures as the fundamental units of speech production, where increased overlap—due to faster speech rates or prosodic strengthening—leads to more robust closures, as exemplified in English alveolar stops assimilating in clusters like "bad ban" through intensified gestural coupling. This framework underscores how fortition emerges from dynamic adjustments in gesture timing and stiffness, rather than static feature changes.[9]Acoustic Correlates
Fortition manifests acoustically through several measurable properties that enhance the perceptual salience of consonants, distinguishing them from their lenited counterparts. One primary correlate is an increase in voice onset time (VOT) for voiceless stops, where the interval between consonant release and the onset of voicing lengthens, often shifting from short-lag values near 0 ms in lenis forms to long-lag values exceeding 50 ms in fortis realizations. This extension arises from greater laryngeal tension and delayed voicing initiation, serving as a robust cue for tenseness in languages like German, where fortis stops exhibit significantly longer VOT compared to lax ones. Similarly, closure duration prolongs in fortified stops, typically from around 70 ms in lenis contexts to over 110 ms in fortis ones, as observed in Trique obstruents, contributing to a more sustained oral constriction.[10][11][11] Spectral characteristics further differentiate fortified sounds, featuring higher overall intensity and sharper formant transitions due to more abrupt articulatory releases. In fortis stops and affricates, the burst energy shows elevated amplitude, often 5-10 dB greater than in lenis variants, reflecting stronger glottal airflow and supraglottal pressure. Formant transitions into adjacent vowels become steeper, with rapid shifts in F1 and F2 frequencies (e.g., F1 rising more sharply from 500 Hz to 600 Hz), which heightens the auditory contrast and perceptual robustness of the consonant. These spectral enhancements, linked to increased articulatory force, underscore fortition's role in amplifying acoustic prominence without altering core segmental identity.[12][11][12] Perceptual experiments confirm that these acoustic cues drive listeners' identification of fortified consonants as perceptually "stronger" or more tense. For instance, longer VOT and closure duration cue fortis categories in stop contrasts, with listeners categorizing stimuli with VOT >50 ms and durations >100 ms as unaspirated or tense stops rather than voiced lenis forms. Increased burst amplitude and sharp transitions further bias perception toward fortis interpretations, as demonstrated in cross-linguistic studies where such properties enhance discriminability and strength ratings. These cues collectively facilitate robust auditory processing, ensuring fortified sounds are reliably distinguished in noisy or rapid speech contexts.[13][14][11]Types of Fortition
Consonantal Fortition
Consonantal fortition refers to phonological processes that strengthen individual consonants by increasing articulatory stricture or enhancing perceptual salience, typically in strong prosodic positions such as syllable onsets or stressed syllables.[5] These changes contrast with lenition by moving consonants toward greater articulatory effort, often resulting in more abrupt closures or enhanced contrasts.[3] A cross-linguistic survey of 81 languages identifies 31 instances of such fortition, predominantly affecting glides and fricatives in domain-initial or pre-vocalic contexts.[5] Manner fortition involves shifts in the manner of articulation that increase constriction, such as the development of approximants or fricatives into affricates or stops. For instance, in Garifuna, the fricative /ʃ/ strengthens to the affricate [tʃ] in stressed syllable onsets, enhancing the consonant's perceptual distinctiveness.[5] Similarly, in spontaneous Received Pronunciation English, /t/ develops affricate-like qualities.[15] These processes are less common than lenition but occur systematically to resolve marked structures in strong positions.[5] Place fortition strengthens consonants through assimilation to adjacent segments, often via palatalization that leads to affrication or other place adjustments. In Kirundi, non-labial consonants palatalize before /j/, resulting in affricates like /t+j/ → [ʦ] across morpheme boundaries, as in /ja-teek-je/ → [jateːʦe].[16] This repair strategy competes with glide strengthening, where the glide itself hardens, but place fortition targets the consonant's articulation to avoid illicit sequences. In Greek dialects, velar /k/ palatalizes to [ʧ] before front vowels, illustrating how place assimilation fortifies the segment.[17] Voicing fortition manifests as devoicing or aspiration of consonants in prominent positions to boost auditory prominence. Positional markedness constraints favor voiceless obstruents in syllable onsets, driving devoicing as fortition; for example, in Chamicuro, glottal contrasts are neutralized in onsets to favor non-glottal realizations.[3] Aspiration similarly strengthens stops in onset contexts, as seen in languages like English where word-initial voiceless stops acquire aspirated releases ([pʰ], [tʰ], [kʰ]) for greater salience.[5] These voicing adjustments occur in strong environments to counteract potential weakening.[18] Consonant strength hierarchies provide a framework for understanding fortition, often inverting the sonority scale where lower sonority equates to greater strength (e.g., stops > affricates > fricatives > approximants).[19] Fortition moves segments up this hierarchy, decreasing sonority—for instance, from fricative to stop—while lenition proceeds downward.[5] Early formulations, such as those by Lass (1984), incorporate aspiration and affrication into these scales, emphasizing inherent properties like stricture over positional factors alone.[19] This hierarchy underscores fortition's role in maintaining robust consonantal contrasts across languages.[19]Prosodic Fortition
Prosodic fortition refers to the strengthening of consonants triggered by suprasegmental features such as stress, intonation, and rhythmic structure, often resulting in increased articulatory constriction, duration, or manner changes like aspiration or gemination.[20] This process contrasts with segmental fortition by emphasizing how prosodic prominence enhances consonantal realization, typically in positions of higher perceptual salience.[21] Articulatory studies show that such strengthening is cumulative, with effects intensifying across nested prosodic domains from syllables to utterances.[20] Stress-induced fortition commonly manifests as greater oral constriction or aspiration in consonants within stressed syllables, particularly word-initial ones, to heighten prominence. For instance, in English, alveolar nasals like /n/ exhibit increased linguopalatal contact in stressed positions compared to unstressed ones, as measured by electropalatography.[20] In languages like South Greenlandic, stress triggers gemination of preceding consonants, as in /awata-t/ realized as [àwáttat], where the stop strengthens to mark the main-stressed foot.[21] This enhancement often prioritizes qualitative changes, such as fricative affrication or stop closure lengthening, over mere duration increases.[22] Intonational effects promote fortition at phrase boundaries, where consonants at the onset of intonational phrases or higher domains undergo strengthening to signal structural divisions. In Korean, coronal stops and nasals show progressively stronger articulations utterance-initially versus word-initially, with peak contact correlating highly (r² > 0.9) with domain strength.[20] German final fortition similarly aligns fortified syllable edges with prosodic boundaries, evolving to mark morpheme and phrase ends through stop strengthening, distinct from devoicing patterns in related languages like Dutch.[23] Rising intonation contours can amplify this at boundaries, though effects are more consistently tied to domain hierarchy than pitch alone.[24] Rhythm and foot structure drive fortition in strong metrical positions, where consonants in head syllables of feet resist lenition or gain added features to satisfy weight requirements. In Tukang Besi, stress on a foot prompts gemination, as in [topːáŋa], enhancing the coda to bimoraicity.[21] Optimality-theoretic analyses propose constraints like MAIN-TO-WEIGHT, which enforce mora insertion or segmental addition in primary-stressed feet, overriding potential weakening in intervocalic contexts.[21] This metrical conditioning ensures rhythmic stability, with fortition appearing in strong beats across iambic or trochaic systems.[22] Cross-linguistically, prosody often overrides inherent segmental tendencies toward weakness, as seen in patterns where stress blocks lenition in otherwise vulnerable positions. In Northwest Saami, stressed codas fortify via epenthesis or gemination, as in /saavnee+ht/ → [(sáavn.neeht)], preventing reduction despite intervocalic adjacency.[21] Romance languages exhibit phrase-initial fortition in all sampled cases (French, Italian, Spanish, Portuguese, Romanian), with word-initial strengthening in subsets, indicating prosodic prominence as a universal trigger that counters lenition-prone environments.[24] These patterns highlight fortition's role in maintaining perceptual clarity amid rhythmic variation.[20]Specific Processes
Post-nasal Fortition
Post-nasal fortition is a phonological process in which obstruent consonants, particularly voiceless stops, undergo strengthening immediately following a nasal consonant, typically resulting in voicing or prenasalization. This fortition enhances the articulatory and perceptual robustness of the consonant cluster, often transforming sequences like /mp/ into [mb] or prenasalized [ᵐb].[25] In many languages, this process applies obligatorily in nasal-obstruent (NC) contexts, preventing voicelessness after nasals to maintain phonetic compatibility within the cluster.[25] The phonetic mechanism underlying post-nasal fortition involves the aerodynamic properties of nasal airflow, which facilitates voicing in the following obstruent. During the production of a nasal consonant, air leaks through the nasal cavity via a lowered velum, reducing oral air pressure and allowing the vocal folds to vibrate more easily during the closure of a subsequent stop—this is known as the "nasal leak" effect. Additionally, subtle movements of the velum, termed "velar pumping," further aid in sustaining voicing by modulating cavity volume and pressure differences. These factors make prevoicing or full voicing highly likely in post-nasal position, contrasting with the higher oral pressure that typically favors voicelessness elsewhere. Aerodynamic models confirm that even moderate glottal adduction settings can yield complete voicing over typical stop closures of around 100 ms after a nasal.[26] Phonologically, post-nasal fortition is frequently analyzed as a rule of regressive assimilation where voiceless stops acquire voicing from the preceding nasal, formalized in generative terms as [-voice] → [+voice] / [+nasal] _[+stop]. This rule often interacts with nasal place assimilation, strengthening the overall NC cluster. In Bantu languages, such assimilation commonly leads to prenasalized stops, where the nasal component reinforces the obstruent's closure.[25] For instance, in Yao (Bantu P.21), the infinitive prefix /ku-/ combines with an object nasal /n-/ before a root-initial stop, yielding voiced forms like /ku-n-dum-a/ 'to order me' pronounced as [ku-n-dum-a], with the stop fully voiced post-nasally.[25] Historically, post-nasal fortition in Bantu languages traces back to Proto-Bantu, where underlying NC sequences were realized as prenasalized voiced stops, such as *búmb- 'swell' or *gend- 'walk'. Over time, these developed into surface voiced obstruents after nasals through vowel loss in prefixes like *mu-, creating new NC contexts that triggered fortition in many languages. This diachronic strengthening reflects a preference for [+voice] and [-continuant] in post-nasal position, countering lenition tendencies elsewhere in the language family.[25] In some Bantu varieties, the process extends to aspiration of voiceless stops post-nasally, as in Kongo (H.10), where /ku-N-pun-a/ surfaces as [kú-m-phun-á] 'to hit me'.[25] Typologically, post-nasal fortition is prevalent in African languages, especially the Bantu family, where it affects nearly all members through inherited NC complexes and productive morphology. It also occurs in Austronesian languages, often in conjunction with nasal prefixation, though patterns vary; for example, prenasalization strengthens stops in certain Western Austronesian varieties. Overall, the process is cross-linguistically common, appearing in about 15 languages from a sample of 197 in one survey, driven by universal phonetic pressures favoring voicing after nasals.[26][25]Place and Manner Fortition
Place fortition refers to changes in the articulatory place of consonants that result in a more peripheral or marked position, often increasing the degree of constriction and perceptual salience. A prominent example is palatal hardening, where velar stops like /k/ shift to alveolar affricates such as /tʃ/ in the context of front vowels or glides, as observed in various Greek dialects. In Northern Greek varieties like Kozani Greek, this shift occurs through strengthened palatalization triggered by vowel deletion, affecting underlying velars to produce palatal or affricate realizations.[17] This process enhances contrast by adding frication, aligning with fortition's role in strengthening articulatory effort.[17] Manner fortition involves shifts toward greater stricture in consonant production, such as the conversion of fricatives or approximants to stops or affricates without nasal influence. For instance, glides like yod (/j/) may fortify to fricatives in strong positions, as seen in Gallo-Romance where Latin jocu > French jou with yod strengthening to [ʒ] "game".[27] Another case is the progression from fricatives to stops, exemplified in positional strengthening where approximants develop into obstruents to satisfy demands for supralaryngeal place in onsets, as modeled in Chamicuro where glottal-like sounds acquire full place features.[3] These changes are frequently driven by assimilatory processes, either regressive or progressive, where a consonant adopts features from a neighboring sound, leading to increased stricture. Regressive assimilation, for example, can propagate place or manner features backward, resulting in fortition when the target sound gains a more obstructed articulation to match the trigger, as in scenarios where onset requirements enforce feature spreading for perceptual clarity.[3] Progressive assimilation similarly contributes, as seen in yod fortition propagating from a following vowel's palatal features.[27] Diachronically, place and manner fortition appear in sound shifts like variants of Grimm's Law, where positional strength in onsets preserved or enhanced stops against spirantization. In Germanic, plain voiceless stops underwent lenition to fricatives in weak positions, but strong onsets (e.g., after obstruents) resisted this, effectively fortifying manner by maintaining stop quality, as in Latin specio > Old High German spehōn "to see".[27] Such independent strengthening highlights how manner fortition can operate alongside chain shifts to preserve systemic contrasts.[27]Cross-linguistic Examples
In Indo-European Languages
In English, a synchronic example of fortition involves the reversal of t-flapping in emphatic or careful speech, where the intervocalic flap [ɾ]—a lenited realization of /t/ following a stressed vowel—strengthens to an aspirated stop [tʰ]. This process is more frequent in formal registers, such as broadcast speech, where flapping rates drop significantly compared to casual contexts, restoring the canonical voiceless alveolar stop to enhance perceptual clarity.[28] In Germanic languages, fortition appears in historical processes that counteract the voicing effects of Verner's Law through devoicing, particularly in final position. Verner's Law, which voiced voiceless fricatives in Proto-Germanic after unstressed syllables, was later overridden in Gothic by a rule of final devoicing, eliminating voiced alternants and restoring voiceless fricatives. For instance, forms like Gothic mīms (from PIE *méh₂tēr 'mimic' stem) show word-final devoicing. This maintained voicelessness word-finally, distinguishing Gothic from other Germanic branches where voicing persisted.[29] In Romance languages, fortition is evident in the realization of Latin /b/ as a full stop in Spanish clusters, particularly after nasals, contrasting with its default approximant [β] in other positions. This diachronic shift preserved the occlusive quality of Latin intervocalic /b/ in nasal contexts during the transition to Old Spanish, where lenition to [β] occurred elsewhere but was blocked post-nasally to avoid perceptual ambiguity with nasals. For example, Latin *campa 'field' evolved to Spanish *campo with [mb], maintaining the stop articulation for articulatory and perceptual strengthening.[30][31] A historical example of fortition in Greek involves the strengthening of Proto-Indo-European sibilants through optional gemination in connected speech, where initial /s/ becomes /ss/ for prosodic enhancement. Step-by-step, this derives from PIE *swéh₂dʰus 'sweet' > Proto-Greek *hédus > Classical Greek ἥδus hḗdus, but in fluent speech contexts like Homer's oîda d’ hóti sú 'I know that you', the initial /s/ of *sú fortifies to [ssú] via resyllabification and lengthening, increasing consonantal stricture across word boundaries. This process, observed in epic and dialectal texts, aligns with broader patterns of sibilant fortition in Greek phonology.[4]In Non-Indo-European Languages
In Bantu languages, such as Zulu, post-nasal environments trigger prenasalization, where underlying voiceless stops surface as prenasalized voiced stops, enhancing laryngeal contrasts and consonantal prominence. For instance, underlying /mp/ surfaces as [ᵐb], as in the noun class prefix *mu- + *pʰa → [mùᵐbʰà] 'two', reflecting a historical and synchronic process that strengthens articulation after nasals compared to lenition elsewhere. Japanese exhibits manner fortition through gemination, particularly in compound formations and Sino-Japanese vocabulary, where singleton obstruents lengthen to geminates, enhancing duration and stricture. This process, known as sokuon insertion, boosts consonant prominence and is phonologically conditioned by prosodic boundaries, distinguishing it from lenition in intervocalic sites. Acoustic studies confirm that geminate obstruents, including fricatives like [ss], have significantly longer duration than singletons, underscoring the fortitive effect. For example, gemination occurs in lexicalized compounds and mimetic expressions.[32] Austronesian languages like Tagalog demonstrate place assimilation during reduplication, a morphological process that copies initial segments and often adjusts place features for articulatory ease. In verbal reduplication marking aspect with nasal prefixes, coronal /n/ assimilates to velar [ŋ] before back vowels or velars, as in /maŋ-ibig/ 'to love' reduplicating to [maŋi-ŋibig] in certain forms, shifting place to enhance coarticulatory stability. This adjustment exemplifies strengthening in reduplicated prefixes, common in Philippine languages to optimize syllable structure and prosodic weight.[33]Theoretical Approaches
In Generative Phonology
In generative phonology, fortition is modeled as a series of ordered rules that derive surface phonetic forms from underlying representations by enhancing segmental strength, often in prominent positions like word or syllable onsets. A canonical example is the rule strengthening continuants to stops word-initially: [+cons, -son] → [+stop] / # __ , which applies in languages where fricatives or approximants articulate with greater closure to improve perceptual cues at boundaries. This approach emphasizes the transformational derivation, where rules apply sequentially to bundles of distinctive features, reflecting universal principles of phonological markedness and naturalness.[34] The foundational Chomsky-Halle (1968) model, detailed in The Sound Pattern of English, exemplifies fortition through rules like aspiration, which strengthens voiceless stops in onset position before stressed vowels: [+cons, -son, -voice] → [+asp] / # __ V[+stress]. This derives forms such as [pʰæt] from underlying /pæt/ ("pat"), where aspiration interacts with prior stress rules in the linear cycle, producing a more forceful release. For English /r/, surface realizations like the retracted approximant [ɹ̠] in word-initial contexts emerge from an underlying [+cons, +cont] specification via low-level stricture adjustments, though primarily as phonetic detail rather than core phonological alternation.[35] Feature geometry refines these rules by organizing features hierarchically, allowing fortition to operate via spreading or insertion under nodes like aperture or stricture. In such models, strengthening can involve increasing closure in the aperture tier to model manner fortition, as in analyses of stop realizations in strong positions without altering place features, integrating with linear derivations.[30] Linear phonology highlights fortition's role in ordered derivations, where it feeds or bleeds other rules like assimilation. In certain dialects of Persian, such as Sharrezaee, for example, word-initial stopping /v/ → / # __ applies before vowel lengthening, deriving [nũm.bɑ] "baker" from /nɑn.vɑ/ while preserving sonority hierarchies; failure of ordering would yield incorrect surfaces. Such interactions underscore the serial evaluation inherent to generative models.[36]In Optimality Theory
In Optimality Theory (OT), fortition processes are analyzed as the outcome of ranked constraint interactions, where universal markedness constraints favoring perceptually salient or robust articulations in prominent positions outrank faithfulness constraints that preserve underlying forms.[3] This parallel evaluation of candidate outputs ensures that strengthening, such as the promotion of fricatives to stops or the addition of aspiration, emerges holistically without sequential derivations. Markedness constraints target "weak" segments—those with reduced articulatory effort or perceptual prominence—in strong prosodic contexts like onsets or stressed syllables, compelling the selection of fortified variants to optimize well-formedness.[3] Central to these analyses are markedness constraints like *WEAK (or equivalently, M-fort), which penalize weak sounds such as glottal stops or unreleased fricatives in strong positions, outranking input-output faithfulness constraints like IDENT-[F] (preserving features like [place] or [continuant]).[3] For instance, in positional fortition, a constraint such as *HAVE-PLACE (banning placeless glottals in onsets) drives the realization of /ʔap/ as [tap] by enforcing a coronal place of articulation for better salience, with *HAVE-PLACE >> IDENT-[place].[3] Aspiration fortition, a common strengthening of voiceless stops, is similarly captured by a markedness constraint ASPIRATION (assigning a violation to unaspirated word-initial stops), which outranks general FAITH(ULNESS) to yield aspirated outputs.[37] The following tableau illustrates this for an input /sip/ (hypothetical underlying form with unaspirated /p/), where ASPIRATION enforces [sipʰ] as optimal:| Input: /sip/ | ASPIRATION | FAITH |
|---|---|---|
| ☞ [sipʰ] | * | |
| [sip] | *! | |
| [sid] | **! |