BrELT coLAB: Fillers are hardwired in our linguistic identities by Alex Tamulis

You’ve probably taught what ‘conversation fillers’ are, but have you wondered how many ‘uhs’ and ‘likes’ you would need to introduce yourself to a large number of classmates on your first day of class?

Sparked by a discussion that took place in the BrELT community, we have invited Alex Tamulis to talk about fillers, fluency, register, cultural identities, prescriptivism, and the like. Check it out!

If you are curious to read the article that ignited this post, click here.



Alex Tamulis has worked in ELT since 1998, teaching at various language centers. He then worked for MTV Brazil as a translator for 3 years, gaining experience in the field of subtitling, translation and interpretation. He holds a BA in Social Communication from Universidade Metodista. After living in Canada for 6 years, where he took college courses at CCEMS Edmonton, he moved back to Brazil and started taking the Linguistics undergraduate program at Universidade de São Paulo. He also holds CPE and CELTA, by the University of Cambridge. Currently he is an academic consultant for Macmillan Education.


In linguistics, a filler is a sound or word that is spoken in conversation by one interlocutor to signal to others a pause, a hesitation, a transition, or even a placeholder.

They fill a beat in the flow of sound. Fillers fall into the category of formulaic language, and different languages have different characteristic filler sounds. In English, the most common filler sounds are “uh” /ʌ/ and “um” /ʌm/. Among speakers, the fillers “like”, “you know”, “I mean”, “okay”, “so”, “actually”, “basically”, and “right” are among the more prevalent.

Have you ever had to listen to an audio file on Whatsapp? You probably have. What if you had to transcribe that conversation? You would probably wonder what to do with those grunts, coughs, sighs, pauses, mumbled words, hesitation markers, etc. You may think that they aren’t that pleasant to listen back to, but these sounds are probably more common than the words between them.

The good ol’ face-to-face communication is a different beast than what you are doing now (presumably reading this post silently from your computer or smartphone screen.) Some people in the communication business might not think that gestures, nodding, pausing and hesitating are legitimate uses of the language, but if you picture a filler-free spoken situation, it is clear that interlocutors would have a hard time conveying their ideas without them.

Not long ago, I’ve bumped into an article that was shared by a BrELTer on Facebook. It sparked my curiosity, since it promoted the idea of removing “um”, “uh”, “like”, “you know”, “well”, and “right” from one’s vocabulary. Why would you yank them out of your lexicon in the first place? I do feel that they play an important role in spoken language, but the article said otherwise. According to the author, those little buggers are “credibility killers” and “verbal viruses”, hence the need for their obliteration.

This raises a series of concerns, regarding how one sees and promotes language to others, in my opinion. It provides fodder for counterpointing her argument with actual research that’s been done to evaluate how fillers play a discursive role in everyday communication. The issues in her article are as follows:

1) Not citing sources

When you state “researchers say that…”, it is implied that you know who your sources are and you’re going to share them in the references section after you’ve finished your article. If not, how’s that assertion any different than “I think that…”? We are all entitled to state our opinions on any topic, but that would be completely different from linking our stance to research that has been done by other people. If you want to speak your mind, make it clear that the idea is yours, whilst not backed by any scientific research. In fact, this is an actual speech genre: it’s called opinion essay.

It’s much easier to back your point of view when you got some research leverage.

In a series of reaction time experiments, Brennan and Schober (2001) found that hesitation markers were beneficial to comprehension, as listeners were faster to select a target object after a filler was used in the stimulus sentence. Similarly, Fox Tree (2001) showed that “uh” facilitated the speed with which listeners were able to recognize upcoming words. Fraundorf and Watson (2011) show that hesitation markers improve recall whether or not they predict upcoming discourse boundaries and that no such effect results from coughs of equal duration, ruling out a processing time effect. In contrast to the symptom hypothesis, this type of explanation for the usage of hesitation markers has been referred to as the signal hypothesis (de Leeuw, 2007). Still other researchers have pointed out that UM and UH can be used to fulfill various discursive functions (e.g. Swerts, 1998; Rendle-Short, 2004; Tottie, 2014). For example, Swerts (1998) showed that hesitation markers can be used as markers of discourse structure, with hesitation markers occurring more often with stronger discourse breaks than with weaker discourse breaks. Similarly, Tottie (2014) argued that UM and UH can be used as discourse markers, with a similar meaning as the discourse markers “well” and “you know”.

We get to the corollary that extraordinary claims require extraordinary evidence, as Carl Sagan puts it. If you’re going to make a bold statement, you’d better be prepared to substantiate it with solid research.

2) Imposing your linguistic biases on people is useless

Oppression leads to revolution, especially if you’re trying to convince a natural- born questioner like me. Being a questioner doesn’t mean bashing skepticism out of spite. It means that you trudge the metaphysical path of language with caution; you tread it lightly.

Imagine that you are watching Netflix right now. What if I say “Don’t watch any Black Mirror episodes, they’re full of pornography, absurdity, dystopia, bleak scenarios, terror, overall grossness and vulgarity! You must not watch it!” Well, if you said that to me, the first thing that would pop up on my mind would be the urge to watch it, since I’m curious as to why you felt that way towards it; it would stir up my curiosity right off the bat.

The same happens to any curious brain out there: the moment I say that certain linguistic choices are a big no-no, ‘tis the moment it instigates [critical thinking] people to wonder and ponder why it is so. Why are these fillers to be avoided? The answer of the article is simply because “they sound bad”. Well, what’s “bad” anyway? Compared to what? Compared to standard written English?

We have to remind ourselves that language is not written language. Language’s been around for thousands of years (linguists estimate its origins at roughly 100 thousand years ago). Writing is expressing language by letters, in which the content of a linguistic utterance is encoded so that another reader can reconstruct it. The key word here is “reconstruction”; it is not the original utterance per se. I’m writing these sentences now, but my brain devises them first. Therefore, we should see writing as ancillary to spoken language (at least in linguistics, anyway).

If it’s a conversation, a spoken situation, it is totally ok to use these fillers, according to context (some contexts are going to prescribe moderation, which is ok). Using utterances adequately and having that social knowledge is what’s at stake here. I highly recommend looking up “communicative competence” (Widdowson, 1983) in order to find out more about it.

3) Language is heavily linked to cultural identity

We all say things a certain way: our way. This is called idiolect. This idiolect interacts with a sociolect, that is, a variety of language associated to a certain social group. Our language changes, according to where we grew up, went to school, how wealthy (or not) our family was, age, gender, and so many other factors out there (look up William Labov and his book, “Sociolinguistic Patterns”, if you want to know more about it).

Most of the time, using fillers is going to reflect who you are (your linguistic identity). But you are not black and white; you have tons of shades of gray within. You’re a network. You are a multiplicity. You can change the way you speak according to context. You do so either to disguise your membership of, or distance yourself from a particular social group, or to move closer to a group you want to belong

to. Thus, you are a collection of linguistic identities! And so are your fillers, your utterances, noun phrases, verb phrases, wh-movements and the like.

Unfortunately, that is not promoted in the school setting. Most students (and their parents) think that language is just a bunch of fixed, rigid rules; as educational standards decline and pop culture disseminates the slangs of the youth, people tend to think we are turning into a nation of functional illiterates, as Pinker beautifully puts it. The same can be said about these fillers. Their use is to be avoided at all costs, since they sound “bad”. Well, mes amies, let me just point out that if these rules are just conforming to tradition, or to the fact that they’ve been perpetuated by certain writers, they could not be further from the truth. The contradiction begins in the fact that the words “rule” and “grammatical” have very different meanings to a linguist and to a layperson (delving into them would stray us away from focusing on conversational fillers. However, if you want to know more about grammaticality, check out Chapter 12 of Steven Pinker’s “The Language Instinct”).

Would not it be much better if we could describe the contexts in which these fillers are used? Would not it be fairer if we showed people situations where they are completely free to be applied to, versus others where it would be wise to shun them away, instead of overgeneralizing and imposing “our way or the highway”? This topic is relevant in pragmatics – check out Brown and Levinson’s theory of “face-threatening acts”: if you think of fillers as a token, a device that helps mitigate certain face threatening acts toward the listener, whilst protecting their self-image in the process, then you’d take them to a whole new level. For the sake of economy, I won’t go into details here.

4) You could use some language policing

It is ok to recommend linguistic choices, according to context. In other words, you could benefit heaps from analyzing your speech, provided there’s a need to do so. If you’re a public speaker, if you’re immersed in professional situations that require less hesitation and more assertiveness, then by all means, go ahead and police yourself; but that has to be your decision, based on your rationale as to why that situation requires you to adapt a certain way. Never take things at face value, especially with something as subjective as the concept of what “beauty” is, and why this word is “aesthetically” more comely than the other.

By all means, I’m not trying to have a beef with anyone! But the fact that a language maven thinks fillers are uncouth, the hobgoblins of discourse, or non-conducive to numinosity doesn’t peg them as “functionless”; if they didn’t have any purpose on communication and processing effort, they wouldn’t be out there, in thousands of

contexts. Whether they like it or not, their use is syntactically sound and ubiquitous in conversation.

They might look and even sound different in the future. In fact, they will. Languages change, no matter how much we try to control it – there’s only so much control you can put on something nobody owns – and that control can certainly be deemed a form of oppression (and I’m not trying to sound political here, it’s just that reinforcing something to someone with no given justification can be seen that way).


Brumfit, C. (1984). Communicative methodology in Language Teaching: The roles of fluency and accuracy. Cambridge: Cambridge University.

Chomsky, Noam (1966). Cartesian Linguistics: A Chapter in the History of Rationalist Thought. New York: Harper & Row.

McWhorter, J. H. (2001). The power of Babel: A natural history of language. New York: Times Books.

Pinker, S. (1994). The Language Instinct. Harmondworth: Penguin. Pinker, S. (1999). Words and Rules . New York, NY: Harper Perennial.

Widdowson, H. G. (1983). Learning Purpose and Language Use. Oxford University Press.

Variation and change in the use of hesitation markers in Germanic languages. Available on

PS: In order to make my point clearer, here’s a YouTube link for an interview done by Jimmy Kimmel with Emilia Clarke, a British actress. Starting around the 30-second mark, she says “like” more than 7 times, and she also uses “um” and “uh” many times. So does Jimmy Kimmel. Now, are you going to think less of them because they’re using those fillers? She even does an impersonation of the so called “valley girl” accent and she keeps on using the word “like”, meaning that this filler might be ubiquitous in, like, most sociolects.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s