Understanding Speech Fluency

• Speech involves simultaneous and successive muscular movements. Fluency is a dimension of speech, along with voice, articulation, and prosody.

What is Fluency?

• Fluency originates from the Latin word "fluere," meaning "to flow."
• Fluency, in language, is often used interchangeably with proficiency.
• A fluent person speaks a language rapidly and continuously without difficulty.
• Fluency includes syntactic, semantic, phonologic, and pragmatic components.
• Phonologic fluency involves constructing long phonological sequences.
• Syntactic fluency involves constructing complex sentences.
• Semantic fluency involves accessing a large vocabulary.
• Pragmatic fluency refers to effective communication in varied speaking situations.

Dimensions of Fluency

• Fluency is effortless, continuous speech at a rapid rate (Starkweather, 1987).
• Fluency domains are continuity, rate, and effort.
• Speech rhythm is also a dimension of fluency.

Continuity

• Continuity refers to the smooth flow of speech between sounds, words, and phrases.
• Fluent speech lacks interruptions.
• Conversational speech is interrupted.
• Interruptions in typical speakers are called disfluencies, including pauses, revisions, and repetitions.
• Interruptions in individuals with fluency disorders are called dysfluencies, such as sound or syllable repetitions and prolongations.
• Stoppages of any kind can be called disfluencies.

Pauses

• Fluent speech can be interrupted by filled/unfilled pauses.
• Filled pauses: utterances like "um," "uh," "I mean," "you know."
• Unfilled pauses: silences longer than 250 milliseconds.
• Conventional pauses are used by proficient speakers to signal importance.
• Idiosyncratic pauses are used by less proficient speakers due to uncertainty.
• Conventional pauses aid comprehension.
• Idiosyncratic pauses hinder comprehension.
• Pauses occur before grammatically loaded words.
• Pauses occur during spontaneous speech tasks.
• Reading tasks have synchronized breathing with pauses.
• Pauses are common before uncertain, long, and phonetically complex words.
• They appear at clause beginnings.
• Frequency and duration of pauses can be documented.

Repetitions

• Repetitions interrupt the forward flow of speech.
• A person will repeat a sound, syllable, word or phrase.
• Sound and syllable repetitions are under "part-word repetitions".
• Word repetitions involve mono- or multisyllabic words.
• Repetition frequency: calculated as percentage (repeated syllables/words divided by the total syllables/words * 100).

False Starts/Revisions, Parenthetical Remarks and Interjections

• False starts/revisions: Interrupting and restarting or changing the utterance. Example: "This is blue a purple bag."
• Revisions include replacing a word with a new word or inserting a word.
• Grammatical structure changes are also revisions.
• False starts/revisions act as self-repair strategies.
• Parenthetical remarks are meaningful but not appropriate for the sentence. e.g: "I mean, like, you know, well"
• Interjections are extraneous sounds (e.g. "uh," "um").
• Speakers use interjections and parenthetical remarks to fill pauses.
• Similar to repetitions, you can calculate the frequency of false starts, etc as a percentage.

Broken Words/Blocks and Prolongations

• Broken words are within-word disfluencies with a break in smooth speech, fixed articulatory postures, and airflow stoppage.
• Prolongations are undue persistence of a speech sound, only with continuants.
• Voiced prolongations have airflow and voicing.
• Voiceless prolongations involve airflow only.
• Frequency and duration of broken words/prolongations can be calculated.

Rate of Speech

• Speech rate is how fast someone speaks.
• Speech rate factors include mood, environment, linguistics, age, culture, articulation speed.
• Fluent speech has a rapid rate.
• Speech rate can be measured in syllables/second, syllables/minute, or words/minute.
• Adult English speakers average 5-6 syllables/second.
• Normal speech rates vary between 80-180 words per minute.
• Intelligible speech can reach 250-280 words per minute.
• Articulation rate (AR) is utterances (without pauses/disfluencies) divided by articulation time.

Effort

• Effort is the muscular or mental energy used in speech.
• Muscular effort is energy spent moving muscles in speech (respiratory, laryngeal, articulatory).
• Mental effort is the planning of speech utterances.
• Fluent speech requires little effort.
• Muscular effort documented with action potentials from speech muscles
• Mental effort is difficult to document, so indirect measures like reaction time (RT) are used.

Rhythm

• Rhythm has temporal regularity.
• Speech has less rhythm than music, but is rhythmic.
• Rhythm divides utterances into phrases and gives speakers some amount of respite.
• Rhythm facilitates speech.
• Language have different rhythms.
• Stress-timed languages: English, French.
• Syllable-timed languages: Hindi, Telugu.
• Mora-timed languages: Japanese.

The Concept of A Mora

• Words divide into syllables, and syllables further divide into mora.

• "Cut" has one syllable.

• CVC structure: consonant-vowel-consonant.

• Each component of syllable is assigned a moraic value.

• Mora is the unit of syllable weight.

• Prevocalic consonants lacks a moraic value.

• A short/lax vowel has one mora.

• Post-vocalic consonants have one mora.

• "Cut" is two moras.

• In English, utterances require 2 moras to be meaningful.

• Stress-timed languages have stress at regular intervals.

• Syllable-timed languages have syllables produced in equal time intervals.

• Mora-timed languages have moras produced at equal time intervals.

• Quantification of rhythm uses pairwise variability index (PVI).

• PVI measures variability in successive vocalic and intervocalic intervals (Low, 2006).

• Normalized pairwise variability index (nPVI): vocalic intervals.

• Raw pairwise variability index (rPVI): intervocalic intervals.

• Stress-timed languages have high nPVI and rPVI.

• Syllable-timed languages have low nPVI and/or high rPVI.

• Mora-timed languages have low nPVI and rPVI.

• PVI values varies based on consonant and vowel occurrences.

• Stress-timed languages: English, Dutch, German.

• Syllable-timed languages: French, Spanish.

• Mora-timed languages: Japanese.

• Children progress from syllable-timed to mora-timed.

• The native accents happens by applying rhythm of one language over another language.

• Learning language involves learning phonemes and prosodic features like rhythm.

• Suprasegmental aspects give nuance and meaning.

Prosody

• Prosody adds fluency through suprasegmental features and rhythm.
• A speech lacks prosody which sound disfluent.
• Rhythm is fluency dimension.
• Elaborate on some prosodic feature.

Stress

• Stress comes from listener or speaker.
• Stress, to listener, relates to change in loudness
• Stressed syllables sounds louder.
• Stress, to speaker, involves more muscular effort.
• Stress helps listeners segment acoustic events.
• Syntactic level marks phase beginning/end.
• Lexical level helps identify verbs from nouns.
• Pragmatic level clarifies content.
• Classifications include phonemic/word-level, morphological, and sentence-level stress.
• Phonemic/word-level stress has a minimum unit of syllable.
• Identify if a syllable is stressed or unstressed contrast to another syllable
• Minimum of two syllables for word and syllable-level stress.
• Free stress: not governed by lexical, syntactical, or morphological rules
• Bound stress has stress fixed on a particular syllable
• Four kinds of stress analysis: unstressed vs stressed syllable place; equal stress, no stress on all syllables
• Sentence-level stress is no meaning change, but relative importance in sentence stress changes the emphasis.
• Three types: primary (non-emphatic, certain syllables/words for the rules of language); emphatic (highlight it or emphasis); contrastive (particular syllable and two contrasting elements in successive sentences/phrases can be differentiated)
• Perceive stress by loudness (Trager, Smith 1951), syllable length (Fant 1958), and pitch prominence (Bolinger 1958). and rhythm (Lieberman 1960). stress differs for each language
• Cue of stress is increased vowel duration/pitch for the distinctions of the languages

Intonation

• Intonation has variations in fundamental frequency.
• Tone refers to pitch variation at words.
• Categorized as tone language, stress accent language and pitch accent languages by tonality throughout the world
• Identification relies on subjective, and objective methods.
• Layout has a baseline and top line.
• The pitch is determined by the language, speaker, and words.
• Intonations are documented. Some information available for Indian languages
• Universality. Languages have unique features. Varies across the dialects of a language.

Factors Affecting Fluency Development

• Fluency development can be affected by internal and external variables.

Age

• Disfluencies increase between age 4-8 (especially word repetitions)
• Disfluencies tend to decrease with age.
• Also, repetition units were highest in the 3 year old and decreased.

Gender

• Sex has no effect on disfluencies.
• Females have better fluency than males.

Bilingualism

• Frequency higher in non-dominant langauge.

Speech and Language Input

• Nature might affect a child;s fluency
• Vary to amount influence development depends on each individual. Peak around 4 years

Typical Disfluency and Stuttering (Historical Perspective)

• Earliest attempts was made by Johnson and colleagues (1959)

• Compared speech of children who stuttered with a equal number who did not .

• Claim Non-fluencies can occur and no mark difference

• States stuttering develops because of a misdiagnosis by overprotective parents and limited knowledge of the subject

• View as Continuity (parts of continuum). States young children experiment with communicative stresses that creates minor disruptions

• States groups overlap and normal and abnormal groups have no major differences

• Terminologies exist (ex: word disfluencies) used to describe CWS and CWNS. Studies also measure and contrast number, type for SLD

• Results shown in 2 or 3 years repeat interjections, revisions more. Those who suffer stutter 1st section multiple times compared to those that do not.

• Studied disfluencies on 24 young children that stutter and do not, the study highlighted; higher percentage of stutter

• Some terms used to categorize. Part word repit and prolong most. 66%, some type of disfluencies

• Studies of both children who does or do not stuttered done in their home language. 2 groups were SLD, results found in significant numbers.

• Studied 90 children form 2 to 5 year range. Measured in six-category system for ID. Major in SLD

• Research for finding differentiating normal and early stutter has been area of research for decades.

• Zebrowski(1991) study, show significantly high for CWNS. Amrose and Yairi from 2 to 5 recorded. First three of most

• Combined in term and other three categorized. Segment had more, calculated was not different then groups and each others.

Type of Disfluencies

• (savithiri 2011) state that cws produces more sud compared to cwns

• Some studieies had report 2x number. High in cws comparing to cwns 8x They show low percentage

• The type had no observations

• High per 100 compared to cwns vs cws. Limited on one or type. This factor differentiates the 2

• Normal has less then 2 vs cws that has more

• The normal does go through the cycle for a few years as a pattern Development trend

• Do to secondary behavior which will later have intergraded stutter. Are rare in develop children in process it happens.

• Assess by showing if a child as TD and the future.

• The assessment happens with a mix of parental data, clinical and if risk in stutter. Info from parent is need. Proves.

• Not sequential it just overlapping. (questions)

1. (1. (1. ( (how much is normal what makes

• Look for their child. Then ask what. And demonstrate and if they see it.

• Look for that child behavior

• You might hear negative, muscular strains or tensions they show. How the child feels

Those can be measured with interview, observation and assess.

Also language analysis (2007)

• Risk and those are factors

1. excessive high the more likely to oral. Negatively review 3 and 10

Assessment

• By a parental viewpoint, kids rarely view them in those ways.
• Family (positive side helps with it, positive attitude)
• Is to have a conversation