Are babies listening to us already in the womb?

Language development is a topic that fascinates me for years and it all started by observing my oldest daughter slowly responding and developing her linguistic skills by interacting with the outside world. Making funny but understandable errors, learning by trial and error…Now when there is a small baby growing in my belly, which response to touches and voices, I wonder how much it actually hears and how much what it hears affects its future development…

Most of the people and research focuses on the brain of the newborn and the language acquisition after the birth, but brain areas corresponding to the language development appear in the brain already during the 30th week of pregnancy, and as soon as from the 18th week of pregnancy, it can ‘hear’ sounds of your body such as your heartbeat and since the 27th week auditory learning becomes possible…There are studies that show that a fetus in the womb reacts to the mother’s voice, to the music, etc. There are also studies that show that when the baby is exposed to one language during pregnancy and adopted at birth, it can afterward still produce mother language sounds or distinguish some words from that language. All these findings are impressive and bring up an interesting question – how is the prenatal exposure to sounds and languages affecting the later development of the language for the baby? Can we improve kids’ linguistic or musical skills by prenatal exposure to foreign languages, by talking to them or playing music to them? Or is how much we talk to the kid affecting our bonding and connection to the baby? Maybe we cannot answer these questions or not all of them, but we can see a bit more, what is happening in the womb, we can explore the life in the darkness and see how the small brain is a step by step developing and responding to the outer world…

Neurophysiology of the fetus

First, let’s see, what happens with the baby brain which enables it to hear and recognize the language, all of which later leads to the fact that newborns might be soothed by familiar sounds like an adult heartbeat or that they prefer mother’s language to other languages…. The onset of hearing in humans is around 27 weeks of gestational age [1], but already since 18th or 19th week, the baby can respond to sounds of the body such as heartbeat or stomach rumbling [2] as the human cochlea starts to be slowly functioning [3]. That is the time when external auditory inputs start to have an effect on the reorganization of the auditory cortex [4]. To test the onset of hearing, [1] performed an experiment, where they exposed the fetus to pure auditory stimuli of different frequencies (100, 250, 500, 1000, and 3000 Hz) – from 19 to 35 weeks of gestational age (GA). Measuring the responses by loudspeaker on maternal abdomen and by ultrasound, they observed first responses for 500 Hz at 19 weeks of GA, then responses to lower frequencies appeared (At 27 weeks, 96% fetuses were responding to 250 and 500 Hz) and the latest ones were responses to higher frequencies (responses to 1000 and 3000 Hz haven’t been observed till 33 and 35 weeks, respectively). As the baby brain matures, a big decrease (20-30 dB) in the intensity level required to elicit a response was observed. This corresponds to the neurophysiological maturation of the inner ear, which seems to be matured at the end of the 8th month when all relevant neuronal synapses take their place [3].

In the 30th week, brain areas corresponding to language development are developing ([5, 6]). As you maybe know, there is observed a left hemisphere dominance for language processing, which seems to be genetically given and we can observe asymmetric development of the brain already before birth. For example, the temporal plane layer in the left hemisphere is larger than in the right one. We can also observe that the left hemisphere has different sensitiveness for speech already from birth (response was measured by electrical activity – event-related potentials, Mehler). Interestingly, the development of cortical regions that are connected to language is slower in the left hemisphere than in the right one (e.g., right temporal plane develops at the 30th week of GA, but left one 7-10 days later) ([7, 5]). The fact that brain regions for language development are distinct already at birth might be a partial verification for Chomsky [8] hypothesis, that there exists a universal innate grammar, which we just fill in with vocabulary and rules.

This above-mentioned development leads to the fact that after 32 weeks of gestation age, the baby may start recognize vowel sounds from your language [2], which further leads to the findings which we will mention later, that the newborn prefers the native language to other languages [9].

Sound and music in the womb – is the baby partying there?

How can we measure what the baby actually hears? It is easy. You can just place a microphone or hydrophone inside the vagina close to the baby’s head and listen…

You will hear a lot of background noise, body noise and mainly low frequencies around 500-700 Hz, all of which are highly attenuated [10, 11, 12]. Nevertheless, mother voice and external speech will still be clearly emerging above the background noise with well preserved prosodic characteristics. [11, 12] found out that even some phonemes/words are still recognizable (up to 30%). All these speech connected sounds are quite silent as they are attenuated significantly. Although you can hear all (close by) external speech, you will get during your life in the womb most familiar with the mother’s voice, because there is significantly better transmission of maternal voice than of external voices – Querleu found out that there is 20dB attenuation of external voices and only 8 dB of mother voices [11]. How can this be explained? It seems that there are 2 pathways transmitting the voice – one is the same for both maternal and external speech, but the mother’s voice is transmitted also through body tissues and bones which enables kids to get more used to this voice.

 This is closely connected to the findings on the newborns, that mother’s voice has calming effects for < 2-week old kids and is more attractive for the newborn than other female voices. On the contrary, newborns do not show any preference for the voice of the father compared to other male’s voices, even after 4-10 hours of exposure to the father’s voice after delivery. [10a] Spence [13] went even further and found out that infants prefer the maternal low-pass filtered voice, but not the maternal whispered voice.

What is another interesting finding is that external auditory stimuli in general increases motor responses (both long term and startle responses) as well as heart rate responses [10], but the mother speech can decrease fetal heart rate – the most when the speech is in low frequencies and around 70dB. The responses of the fetus were measured for tones, vibratory frequency noises, or EAL stimulations during quiet and active sleep and during active and quiet wakefulness of the baby – it has been found that although to tones the baby starts to respond at 27-28th week of GA, to EAL stimuli motor responses appear already in the 24th week [10]. Bigger responses can be observed during wakefulness than during sleep. Unborn babies might even start to cry if exposed to a sudden loud noise.

And how is it with the music? Feijoo [14] measured kids at delivery and 4-5 days old and found out that kids were soothed by the music which their mother listened regularly during last 3 months of the pregnancy way more than any other music.

Postnatal effects of prenatal stimuli

So you can see that it has been found out, that familiar sounds might soothe crying newborn [10], including adult heartbeat, mother’s voice[13, 10], music [14], or soap opera theme [15].

Question is, how can we find out what a newborn actually prefers? There is an experiment called the non-nutritive sucking behavior of the newborn which is supposed to measure babies’ responses to individual stimuli. What should this mean? It is easy, this technical term hides an easy thing – behavior while sucking on a pacifier. So basically we measure how much the baby is sucking, what are the pauses between sucking and frequency of sucking. So easy :).

Using this method, it has been found that newborn babies exposed to the adult heartbeat (72 beats/s) were sleeping earlier and gained weight quicker [16]. Also, intrauterine background noise had a short term pacifying effect ([10a], Murooka [17]). Familiar stimuli were also preferred by a child in selection tasks.

Back to mother’s voice…DeCasper [18] found out that 2-3 days old babies preferred a story that mother read 6 weeks prior to pregnancy to the one they never heard. And have you ever heard about fetal ‘soap’ addiction? It was described by Hepper in 1988 [15], [19]. What they found out? That there were mothers who regularly watched a soap opera ‘Neighbors’ during the last weeks of pregnancy (some even twice a day). Maybe you can already guess what they observed for these babies after delivery. A crying baby who was ‘listening’ to the soap opera already in the womb, is more likely to stop crying when the program begins than the baby of a mother who was not watching it in her pregnancy.

So far I was mainly speaking about voice and sounds, but what about the actual languages? Moon [9a], measured 40 infants about 30 hours old in the nursery using the test for sucking behavior when listening to vowels from native and unfamiliar language. They found out that babies were sucking longer for the foreign language than for the native tongue (independent on time since birth, 7-75h), which indicates that they can differentiate between these two languages and it seems that the ambient language to which fetuses are exposed in the womb affects their perception of the native language at phonetic level. In another study, Moon et al. [20] measured 2 day old kids of French and Russian mothers while the same bilingual woman was talking to them. These kids again preferred the mother’s tongue to the unfamiliar one and the same behavior was observed also for low-pass filtered versions, which kept only the prosodic cues of the languages. On top when they measured kids from families speaking neither French nor Russian, they were not able to discriminate between these 2 languages. [21]

Maybe you might be a bit skeptical about this sucking behavior test…Then there is another study conducted by May et al. [22], who measured a neural activity of the brain using NIRS (near-infrared spectroscopy) study on 20 monolingual English 0-3 days old neonates. She let them be exposed to low-pass filtered sentences of forward English (familiar language), forward Tagalog (unfamiliar language), and backward English and Tagalog (non-language) and measured neural activation by NIRS. They found out bilateral (in both hemispheres) activation connected to the familiarity of the language and observed different activation for Tagalog (unfamiliar language). This indicates that exposure to the language in uterus influences the brain responses of the baby to the native language after birth.

So it actually seems that babies in the womb really listen to us as soon as their hearing enables them so, although what comes to them is highly attenuated and filtered for high frequencies. After several months of this auditory experience, they become attracted to the familiar sounds and patterns of the native language(s). But it is possible that they actually learn even more from their prenatal experience as they might become familiar with the prosody of the languages heard in utero (‘prenatal prosodic bootstrapping’), which might be crucial for later language acquisition [23]. To conclude, we can see that long-term premature exposure to auditory (or other sensory stimuli) in utero might lead to both prenatal and postnatal responses to these stimuli.

And one last thing to mention, the auditory and language system as a whole continues, of course, to develop also after a birth – e.g., hearing develops for another 5-6 months after birth and language areas keep developing for years, so it needs continuous stimulation through speech, music, and other sounds, otherwise we might miss the important times (so-called sensible periods) for its development which might lead to its underdevelopment, as observed for linguistically or socially deprived ‘wild kids’…but that is a topic for some other time…


  • [1] Hepper PG, Shahidullah BS (1994) Development of fetal hearing. Arch Dis Child 71: F81–F87. 
  • [2] 
  • [3] Pujol, Rémy, and Adini Uziel. “Auditory development: peripheral aspects.” Handbook of human growth and developmental biology 1.Part B (1988): 109-130. 
  • [4] Chang, Edward F., and Michael M. Merzenich. “Environmental noise retards auditory cortical development.” science 300.5618 (2003): 498-502. 
  • [5] Gazzaniga, Michael S. The new cognitive neurosciences. MIT press, 2000. 
  • [6] Department of Biochemistry and Molecular Biophysics Thomas Jessell, Steven Siegelbaum, and A. J. Hudspeth. Principles of neural science. Eds. Eric R. Kandel, James H. Schwartz, and Thomas M. Jessell. Vol. 4. New York: McGraw-hill, 2000. 
  • [7] Chi, Je G., Elizabeth C. Dooling, and Floyd H. Gilles. “Gyral development of the human brain.” Annals of Neurology: Official Journal of the American Neurological Association and the Child Neurology Society 1.1 (1977): 86-93. 
  • [8] Chomsky, Noam. “Recent contributions to the theory of innate ideas.” A Portrait of Twenty-five Years. Springer, Dordrecht, 1967. 31-40. 
  • [9] Moon, C., Lagercrantz, H., & Kuhl, P. K. ( Language experienced in utero affects vowel perception after birth: A two-country study. Acta Pediatrica, 102, 156-160. Click here to receive a reprint 
  • [10] Lecanuet, Jean-Pierre, and Benoist Schaal. “Fetal sensory competencies.” European Journal of Obstetrics & Gynecology and Reproductive Biology 68 (1996): 1-23. 
  • [11] Querleu, Denis, et al. “Fetal hearing.” European Journal of Obstetrics & Gynecology and Reproductive Biology 28.3 (1988): 191-212. [12] Busnel, Marie-Claire, C. Granier‐Deferre, and J. P. Lecanuet. “Fetal audition.” Annals of the New York Academy of Sciences 662.1 (1992): 118-134. 
  • [13] Spence, Melanie J., and Mark S. Freeman. “Newborn infants prefer the maternal low-pass filtered voice, but not the maternal whispered voice.” Infant Behavior and Development 19.2 (1996): 199-212. 
  • [14] Feijoo, J. “Le foetus. Pierre et le Loup.” L’aube des sens (1981): 192-209. 
  • [15] Hepper, Peter G. “Fetal” soap” addiction.” Lancet (London, England) 1.8598 (1988): 1347-1348. 
  • [16] Salk – The effects of the normal heartbeat sound on the behavior of the newborn infant 
  • [17] Murooka, Hajima. “Analyse des sons intra-utérins et leurs effets tranquillisants sur le nouveau-né.” (1976). 
  • [18] DeCasper, Anthony J., and Melanie J. Spence. “Prenatal maternal speech influences newborns’ perception of speech sounds.” Infant behavior and Development 9.2 (1986): 133-150. 
  • [19] 
  • [20] Fifer, William P., and Christine Moon. “Psychobiology of newborn auditory preferences.” Seminars in perinatology. Vol. 13. No. 5. 1989. 
  • [21] Mehler, Jacques, et al. “A precursor of language acquisition in young infants.” Cognition 29.2 (1988): 143-178. 
  • [22] May, Lillian, et al. “Language and the newborn brain: does prenatal language experience shape the neonate neural response to speech?.” Frontiers in Psychology 2 (2011): 222. 
  • [23] Gervain, Judit. “The role of prenatal experience in language development.” Current opinion in behavioral sciences 21 (2018): 62-67.
  • [WN] 
  • [24]

How sleep during pregnancy affects a fetus and its sleep?

During all my pregnancies I always wondered how are fetus sleep/non-sleep periods related to my own sleep. And how my own sleep is related to the baby brain development and affecting the kid in general. Now it is my third pregnancy. I have way more work as a researcher, two kids already occupying me during the day and waking me to prepare them breakfast in the morning (somehow always too early) and there are too many nights when I just get too little sleep. It seems as the proper time to finally explore the effects of my behavior in detail and try to understand this hidden secret which is happening somewhere inside my growing belly.  Is it anyhow harmful for the kid when I do not sleep enough as it still can happily sleep its 90% of time, or is my tiredness affecting the kid’s development by different substances floating to the womb or via changes in hormone levels in my body? Does the kid’s brain develops differently during my sleep and during my awakened periods? Another interesting thing for me was that the small baby in my womb seemed to move more when I was resting and its movements seemed almost unnoticeable when I was hiking or being anyhow active. When and how it actually sleeps and how it might feel?

How sleep deprivation affects adults is quite known and there are many resources on this topic. It has been shown that prolonged sleep restriction (less than 6 hours per night) or prolonged extended sleep duration (more than 9 hours of sleep per night) are associated with increased mortality [1]. But how it affects the fetus, is way more hidden secret. So let’s dwell to this fascinating topic…

Fetus sleep

Sometimes we think about a fetus and newborn as two totally separate beings. Or at least for me, the kid which is still in my womb is totally different from the newborn which will come out and start being hearable and touchable. There are those moments, which make the unborn baby more present for me – like the ultrasonic or when you feel the baby kicking. Anyway, at least for me, the baby in the womb is still very silent creature, which is difficult to understand. Sometimes I have a feeling that it is just the belly, sickness, tiredness and so on, which I have to deal with, forgetting, that inside is a new small human being growing. The current neuroscience research reveals more and more about what is happening inside the womb during the pregnancy. And it is indeed fascinating story. As measuring the fetus brain activity in the womb is very challenging (there exist fetus fMRI [2], or fetal EEG measurements from scalp of a baby during labor or from mother abdomen [3], but these measurements are very prone to noise and the interpretation of the data is very challenging), we are mainly restricted to external explorations. It seems that all the way long from the week first till the delivery, there is a continuous process of cognitive development of the kid – not only preparing the physical body, body and brain cells, not only developing digestion, training breathing and motorics, but also improving voice recognition, language development, learning and memory, other cognitive functions, and maybe even dreaming. This process continuously follows to a newborn stage. There is no strict border between a fetus and a newborn.  

And same it is with the sleep. The baby in the womb spends almost all the time asleep. As it approaches 32 weeks, it still sleeps 90-95% of the day. Every week the amount of sleep is a bit reduced and at delivery time, they sleep around 85-90% of the day, which is the same time, as the newborn will sleep when coming out of the womb. It might be, that during sleep the fetus is even dreaming as it spends some time in REM phase of the sleep with eyes moving quickly back and forth. “Around the seventh month of a fetus’ development, the first rapid eye movements are seen. The brain of the developing embryo appears to cycle every 20 to 40 minutes between REM sleep, in which brain activity rivals that of consciousness, and non-REM sleep, in which the brain rests.”[4] Swab [5] explored the fetus of a sheep and came to the conclusion that “Sleep does not suddenly evolve from a resting brain. Sleep and sleep state changes are active regulated processes,” How are these different stages of the brain activity important for the brain development is still a research question. “Cyclic changes in the activity of neurons, for instance, may stimulate the other nerve cells to find and connect with each other to set up complex networks in the brain.”[4]

You can see visualisation of fetus sleep cycles by 26th to 30th week of pregnancy on this video:

It is known that sleep and its individual stages are important for fetal and early neonatal brain development, including development of the neurosensory and motor systems, for learning and memory in the fetus and neonate and for preserving the brain plasticity.[6]That is why the brain of an infant or child is more active during some stages of sleep than during wakefulness and awake activities. [7] The importance of sleep and sleep cycles for the fetus in utero and the preterm infant in intensive care unit is known for more than 40 years. [6] REM sleep deprivation in kids might result in disordered sensory system, disordered learning and memory, disordered or disrupted learning and memory creation, loss of cortical plasticity into adult life and also smaller adult brain size. The deprivation of Non REM sleep might result in decreased learning and memory consolidation from sensory experiences, less ability to learn in childhood and adult life and loss of brain plasticity into adult life. [7]

The main question for me is how my behavior is actually affecting the sleep of the fetus? It seems that when I am more active, the kid can more happily sleep and when I am resting, it starts to wake up. [8]. In [9] Kintraia explored the daily rhythmicity of 2500 fetuses using the electrotachogram and they found out that “daily rhythmicity of fetal electrocardiogram resembles that of the mother; however, the phase of the rhythm is opposite to that of the mother. “Active” (from 9 a.m. to 2 p.m. and from 7 p.m. to 4 a.m.) and “quiet” (from 4 a.m. to 9 a.m. and from 2 p.m. to 7 p.m.) periods of activity were identified.” A baby may be more active about an hour after the mother eats. This is because of the increase in sugar (glucose) in the mother’s blood. Fetal movement normally increases during the day with peak activity late at night.”[10]

Anyway, there is evidence that the mother activity might directly affect the fetus behavior “ultrasounds have shown a fetus bouncing up and down when the mother laughs.” [11] Also mother movements and activity actually might make it more pleasant for the kid to sleep then when she is in a calm position. What is interesting is that the baby’s brain won’t produce melatonin till the age of 3 months to develop its own circadian clocks. So in the womb it relies on signals from your body clock as melatonin can cross the placenta and after birth, some amount of melatonin can still be passed to the baby through breastmilk. Until circadian clock develops, newborn sleep patterns will be fairly irregular. [12] It is also interesting to see the research of the fetus activity based on the mother sleep position. In [13] they explored 36 pregnant women and found out that when the mother slept on her back, the fetus was less active. As the fetal activity is one measure of its well-being, it is suggested for women to sleep on their side during the later pregnancy (after 28 weeks) to reduce the risk of stillbirth.

You maybe would like to predict the sleep pattern of the newborn and other infant behavior from the fetus movements. Unfortunately, based on the DiPietro study on 52 healthy fetuses, there is no evidence for such predictions [14, 15],

How mothers sleep deprivation affects the fetus?

It might not be only due to the fact that you have a lot to do in the evening why you cannot have a really long night of well-quality sleep. I have to say that in this sense I was quite lucky as I sleep typically quite well. Maybe due to the fact that I typically go to bed later than I should, I have no problems to fall asleep directly and if I do not drink a liter of water before sleep, I can survive without toilet sometimes even till the alarm or kids wake me up. So my main problem is to find enough hours for sleep. Many women however have other reasons for sleep deprivation – insomnia, restless legs syndrome, frequent nighttime urination, etc.[16, 17] My concern is now about my night papers deadlines, when I do not get enough sleep, what then happens to the baby?

“Sleep deprivation during pregnancy has been associated with longer labor, elevated perception of pain and discomfort during labor, higher cesarean rates, preterm labor, and higher levels of pro-inflammatory serum cytokines. Limited research has also indicated a possible relationship between sleep deprivation and preterm births and postpartum depressive mood.” [18] However the studies are typically done on poorly controlled studies and non-representative samples. Study on more than 1000 women showed that Women with severe snoring were at high risk for low birth weight, and fetal-growth-restricted neonates  and women with sleep deprivation (≤5 hours sleep) were at high risk for preterm births, with the highest risk observed for medically indicated preterm births [19,20].

The most interesting topic for me is if the mother sleep deprivation might actually also affect emotional and cognitive development of the baby. One of the effects on the cognitive development of the newborn might be the fact that the sleep deprivation during pregnancy increases the risk of postpartum depression, which in consequence increases the risk that the mother-child bonding is not created properly which in consequence can affect the kids emotional and cognitive development. Corresponding to the more direct effects on the cognitive development of the child, it is very challenging to perform the real experiment on humans as it is not really ethical to sleep deprive women and create meaningful controlled studies.

At least, there have been done several interesting studies using rats. In [21] they sleep deprived pregnant rats and then investigated the offsprings and found impaired hippocampal-dependent spatial learning and memory, and increased depressive- and anxiety-like behaviors. In [22] they observed “increased brain derived neurotrophic factor (BDNF) in the hippocampus of fetuses of sleep deprived mother rats, which indicates that sleep restriction of mothers during the final week of gestation (last trimester of rats pregnancy) may affect neuronal growth factors in a fetal brain structure, in which active neurogenesis occurs during the deprivation period. However, no changes in the total reactive oxygen species (ROS) in the cortex, hippocampus, or cerebellum of the fetuses were detected.” In [23] they found out that sleep deprivation during late pregnancy can produce hyperactivity and increased risk-taking behavior in rat offspring.


YouTube – New video (Specify problems by NL and demonstration)

Specifying dual-arm robot planning problems by natural language and demonstration

New video corresponding to our RAL and ICRA 2019 publication is now available on our YouTube channel

Behrens J., Štěpánová K., R. Lange, R. Skoviera. “Specifying Dual-Arm Robot Planning Problems Through natural language and Demonstration.” RA-L and ICRA (2019)


Paper support materials:…