How Esophageal Speech Works After Laryngectomy

Esophageal speech is a method of speaking without a voice box, used by people who have had their larynx surgically removed. Instead of vibrating vocal cords, the speaker swallows or injects air into the upper esophagus and then releases it, causing a small segment of tissue to vibrate and produce sound. The technique is hands-free and requires no device, but it produces a distinctly low-pitched, rough voice that takes months of practice to develop. Once the dominant form of post-laryngectomy voice rehabilitation, esophageal speech has declined sharply over the past four decades as prosthetic alternatives gained ground, though recent interest in it is quietly growing again.

How the Voice Is Actually Produced

After a total laryngectomy, the windpipe is permanently rerouted to a hole in the neck called a stoma, and the connection between the throat and the airway is severed. This means exhaled air from the lungs no longer passes through the throat at all. Esophageal speech works around that problem by using a completely different air source: the esophagus itself.

The speaker traps a small pocket of air in the top of the esophagus, then pushes it back up. As the air passes through a band of muscle and tissue at the junction of the pharynx and esophagus, that tissue vibrates. This structure, called the pharyngoesophageal (PE) segment, acts as the substitute voice box. The vibration it produces is shaped into words by the tongue, lips, and jaw just as in normal speech. Videofluoroscopy studies comparing this mechanism across speech modes have found that the PE segment sits in roughly the same location and stretches to a similar length whether the air comes from the esophagus or through a surgical voice prosthesis, though the prosthesis-driven version tends to produce more effective-sounding speech overall.¹

There are two main ways to get air into the esophagus. In the “injection” method, the speaker uses tongue and throat movements to pump air downward, somewhat like a reverse swallow. In the “inhalation” method, the speaker creates negative pressure in the esophagus by expanding the chest, drawing air in through the mouth and down. Most speech therapists teach both, and skilled speakers often blend them depending on what they are saying.

What It Sounds Like

Esophageal speech is unmistakably different from a normal voice. The PE segment vibrates at a much lower frequency than healthy vocal cords, giving the voice a deep, gravelly quality. Measurements comparing esophageal, tracheoesophageal, and normal speakers have found that esophageal speech has the lowest fundamental frequency of the three groups. Tracheoesophageal speakers, who use lung air redirected through a prosthesis, averaged a fundamental frequency roughly 25 Hz higher than esophageal speakers.²

Volume is another limitation. Because the esophagus holds far less air than the lungs, each “charge” of air lasts only a second or two. Speakers produce short bursts of sound and must frequently re-inject air, which creates noticeable pauses and limits how loudly they can project. Acoustic studies show that tracheoesophageal speech is more intense than both normal and esophageal speech, while esophageal speech tends to be the quietest of the three.³ Background noise in a restaurant or on a busy street can easily drown out an esophageal speaker.

Despite these acoustic constraints, tracheoesophageal speech and esophageal speech are distinct in measurable ways. Discriminant analysis of acoustic and timing features has perfectly categorized speakers into their respective groups, confirming that the three methods of speech production are acoustically and temporally distinct from one another.⁴ In plain terms, a listener can tell the difference between someone using esophageal speech, someone using a voice prosthesis, and someone with a natural larynx.

Intelligibility and How Well Listeners Understand It

The roughness and low volume of esophageal speech raise an obvious question: can people actually understand it? The answer is yes, but with caveats. When experienced and inexperienced listeners rated recorded samples of esophageal and tracheoesophageal speakers reading aloud and speaking freely, both groups of judges ranked the speakers similarly. The tracheoesophageal speakers were rated as more intelligible overall, but the gap between the two was not so large that esophageal speakers were unintelligible. Interestingly, both listener groups and both speaking tasks (reading versus conversation) produced consistent rankings, suggesting that intelligibility differences are real and not just an artifact of listener familiarity.⁵

Intelligibility varies enormously from person to person. Some esophageal speakers develop clear, conversational speech that strangers can follow without much difficulty. Others remain hard to understand even after extensive training. The skill ceiling is high, but so is the floor, and the range between the best and worst speakers is wider for esophageal speech than for prosthesis users who have a more consistent air supply to work with.

How Many People Successfully Learn It

Success rates in the published literature are all over the map, in part because researchers have never agreed on what “success” means. Some studies count anyone who can produce a few words; others require fluent conversational speech. A review of the evidence found reported success rates ranging from as low as 25% to as high as 86%, depending on the criteria used and the population studied.⁶ That range is so wide it is almost meaningless on its own, but the takeaway is that a substantial minority of people who attempt esophageal speech never achieve functional communication with it.

What predicts failure? Surgery-related factors like the extent of tissue removed and whether the patient received radiation therapy play a role. Psychological readiness and motivation matter too. One hypothesis that has been tested is whether esophageal motility, the coordinated muscle contractions that normally move food down the esophagus, differs between people who learn esophageal speech and those who do not. A study comparing the two groups found that most motility measures were the same, but people who successfully acquired esophageal speech actually had lower esophageal contraction strength at a specific measurement point.⁷ The authors speculated that a more relaxed esophagus may make it easier to trap and release air. This is a single finding and far from definitive, but it hints that the physical properties of each person’s esophagus contribute to whether the technique clicks.

Another commonly suspected culprit is acid reflux. It seems intuitive that stomach acid washing up into the esophagus would irritate the PE segment and hamper speech. But when researchers measured reflux directly with pH probes and compared proficient and nonproficient esophageal speakers, they found no meaningful difference between the two groups. Reflux does not appear to be a major barrier to learning esophageal speech.⁸

When the PE Segment Does Not Cooperate

Even with good technique and motivation, the PE segment itself can be the problem. In some patients, the muscles around this area go into spasm, clamping down too tightly for air to pass through and vibrate the tissue. This condition, called pharyngoesophageal spasm, can make both esophageal and tracheoesophageal speech difficult or impossible. It is one of the more frustrating post-laryngectomy complications because the patient may be doing everything right and still producing no usable voice.

Botulinum toxin injections into the PE segment have emerged as a treatment for this spasm. In a study of 43 patients treated for PE dysfunction with botulinum toxin A, 86% reported subjective improvement in their symptoms, with over a third improving in both swallowing and voice. The injections were delivered by several different guidance methods, and no significant complications were reported.⁹ A separate study focused on patients using voice prostheses found that 94% reported better voice quality after botulinum toxin injection, with the effect lasting an average of about 20 weeks before repeat treatment was needed.¹⁰ The treatment is not permanent, but it can make the difference between having a usable voice and having none.

How It Compares to Other Post-Laryngectomy Options

People who lose their larynx generally have three paths back to speech: esophageal speech, a tracheoesophageal voice prosthesis (TEP), or an electrolarynx. Each has trade-offs that matter in daily life.

A TEP involves a small one-way valve surgically placed between the windpipe and the esophagus. The patient covers the stoma with a finger or a hands-free valve, and exhaled lung air is diverted through the prosthesis into the esophagus, vibrating the PE segment. Because lung air provides far more volume and duration than the small pocket of air used in esophageal speech, TEP speech is louder, longer per breath, and acoustically closer to a normal voice.¹¹ It is also rated as more intelligible by listeners.¹² The prosthesis does require ongoing maintenance and periodic replacement, and it can develop leakage or fungal overgrowth, but it has become the most widely used voice rehabilitation method over the past two decades.¹³

An electrolarynx is a handheld device pressed against the neck or cheek that generates a buzzing vibration. The user shapes the buzz into words with the mouth. It is the easiest method to learn and can be used almost immediately after surgery, but the resulting voice sounds robotic and mechanical. It also ties up one hand during use.

Esophageal speech’s advantages are simplicity and independence. There is no device to maintain, charge, or replace, and both hands stay free. For someone who achieves proficiency, it can feel like the most natural of the three options because it relies entirely on the body. Its disadvantages are the long learning curve, the low volume and short phrasing, and the wide gap between the best and worst outcomes.

The Decline of Esophageal Speech and Signs of a Comeback

Esophageal speech was once the gold standard for post-laryngectomy communication. For decades after the technique was refined in the early twentieth century, it was the primary goal of speech rehabilitation. That changed dramatically with the spread of tracheoesophageal puncture procedures starting in the 1980s. The acquisition and use of esophageal speech has decreased significantly over the past 40 years, a decline that closely tracks the rise of TEP voice restoration.¹⁴

Yet the picture is not one of total replacement. Some patients cannot have a TEP placed, or their prosthesis fails repeatedly, or they simply prefer not to depend on a device. In parts of the world where prostheses are expensive or hard to obtain, esophageal speech remains the primary rehabilitation option. The same 2022 paper that documented the decline also raised the question of whether esophageal speech is returning as a viable option, reflecting growing interest in situations where TEP is not ideal. In Japan, patient associations still run structured esophageal speech training programs, with experienced speakers serving as peer trainers.¹⁵

Quality of Life and the Emotional Side

Losing your voice is not just a communication problem. It reshapes social interactions, professional identity, and self-image. A systematic review of quality-of-life outcomes for esophageal speakers found a mixed picture: some patients reported improved voice-related quality of life and better scores on standardized voice impairment measures, while others found the technique too difficult or the resulting voice too unsatisfying to improve their daily experience. Results were not uniformly positive, with a subset of patients reporting minimal improvement.¹⁶

When esophageal speakers were compared to users of a pneumatic artificial larynx on quality-of-life scales, there was no significant difference in social-emotional or physical functioning scores, suggesting that neither method clearly leads to a better or worse life experience than the other. Among esophageal speakers specifically, time since surgery had a significant effect on physical functioning and overall scores, meaning that quality of life tended to change as people adapted over the months and years following their operation.¹⁷

Gender Perception and the Low-Pitched Voice

Because both esophageal and tracheoesophageal speech produce a low, hoarse voice, women who use these methods face a specific social challenge. In normal speech, pitch is one of the strongest cues listeners use to identify gender. When that cue is stripped away, women are at risk of being perceived differently. A study of tracheoesophageal speakers found that while female speakers were still correctly identified as female by listeners, they were rated as more masculine or gender-neutral on a femininity scale compared to male speakers, who were rated comfortably within the masculine range.¹⁸ This effect is likely at least as pronounced for esophageal speakers, whose fundamental frequency is even lower.

For women adjusting to post-laryngectomy life, this perception gap can add another layer of distress on top of everything else. It affects phone conversations especially, where visual cues are absent and voice is the only information a listener has. Some speech therapists work with female patients on strategies to modulate other vocal qualities like intonation patterns and speech rate to convey femininity through channels other than pitch.

Challenges in Tonal Languages

Languages like Mandarin, Cantonese, and Taiwanese rely on pitch contours to distinguish word meanings. A syllable said with a rising tone can mean something entirely different from the same syllable said with a falling tone. This creates a particular challenge for esophageal speakers, who have limited control over pitch.

Acoustic analysis of Taiwanese esophageal speakers found that while the overall shape of the pitch contour (whether it rises, falls, or stays flat) did not differ significantly from normal speakers, the starting pitch was different, and the amplitude was significantly lower than in normal speech. Esophageal speakers produced the quietest speech of the three groups tested, which included normal laryngeal speakers and pneumatic artificial larynx users.¹⁹ The preserved slope of the pitch contour is encouraging: it suggests that esophageal speakers can, to some extent, produce the tonal patterns their language requires. But the reduced volume and altered starting pitch mean listeners may still struggle, particularly in noisy environments or over the phone.

AI-Driven Voice Restoration

One of the most intriguing developments in the field is the use of artificial intelligence to transform esophageal or mechanical speech into something that sounds like the patient’s original voice. A recently published framework described an AI-driven neural voice conversion system designed to take the rough, low-frequency output of a laryngectomee’s speech and map it onto recordings of the patient’s preoperative voice.²⁰ The goal is not just improved clarity but restoration of acoustic identity, the specific vocal characteristics that make a person sound like themselves.

This technology is still in early stages, and the published work to date describes a technical framework rather than large-scale clinical results. But the concept addresses one of the deepest losses patients experience: not just the ability to speak, but the loss of a voice that family, friends, and colleagues recognize. If voice conversion tools become practical and affordable, they could change the calculus for people choosing between rehabilitation methods. Esophageal speech, which is free and device-independent, could become the input signal for an AI system running on a smartphone, combining the simplicity of the old technique with the vocal quality of something entirely new.