Research background and interests
“Becoming a researcher was not always my dream as a kid. In fact, I wanted to become many things – from priest to military. Even when I was 17 years old I was still very focused on studying biology and genetics - I had chosen all the electives in biology in my high-school back in Chile. But then one day a new professor came to teach Spanish. He was very young, fresh from university, which might explain his enthusiasm and motivation to offer a new elective in Latin. I took this elective and became really interested in grammar. Coincidentally I was already familiar with Chomsky’s political work, and I remembered that he had written about language as well. I started to read this work, and it really blew my mind. In fact, it completely changed my plans for the future – three months before I did the final tests, I switched all my tests in biology to electives in humanities, and decided to continue with a BA in Linguistics, expecting that I would be taught Generative Grammar.”

“However, the Bachelor’s programme at the Pontifica Universidad Católica de Chile was a degree in Literature & Linguistics. The main focus of this programme was literature, and the linguistics part only educated us to become language teachers. This was not at all what I had expected of linguistics. I wanted to do experimental work, and tried to incorporate this in the programme as much as I could through projects and internships. Maria Cristina Arancibia, a new teacher who had done her PhD recently, helped me to do my first psycholinguistic research, which is how I got into contact with the school of psychology. There I met a professor Diego Cosmelli who had just finished his Post-Doc in France, and he was building an EEG lab – literally with hammers and nails. To do EEG’s you need a Faraday cage, which is an electromagnetically sealed cage within a cage. And this was exactly what turned out to be my first job in research – building the Faraday cage. From then on they started to train me there, because I showed interest in research, and the only thing I knew about linguistics was how to teach English as a foreign language. I got involved in recruiting participants, designing stimuli, and then finally they wanted to design an experiment that involved language. As I had the official title of a Linguist, they put me on this project, and there I found that I had a knack for experimental design.“

Coming to Amsterdam
“This experience in experimental design was what got me here in Amsterdam. As I had graduated from my Bachelor with relatively good grades and had hands-on experience with experimental work, I got a scholarship to study abroad. This helped me to pay the master’s programme in Brain & Cognitive Science at the University of Amsterdam, which was very expensive for non-Europeans – approximately a €1000,- a month. I enjoyed this programme very much, as I was finally learning about the things I had been interested in since highschool, although there was only one specific course in linguistics. But I learned a lot about Cognitive Science in general, always keeping in mind how this related to what I am most interested in scientifically: language, and how it is processed unconsciously.”

“This is also what I focused on in two of my research projects in the programme, and a theoretical piece I wrote: I looked at statistical implicit learning, then I looked into theories of language processing and of grammatical processing specifically. Here I realized that both implicit learning and grammatical processing were automatic, unconscious processes that occur very early after stimulus processing. I mixed that theory with an experimental paradigm called “masked priming” in an internship with Judith Rispens. I met Judith Rispens in the course Topics in Language and Cognition. This was the first time that I met a linguist who was interested in psycholinguistics and so I contacted her. That is how we designed this internship project in which I got the chance to design my own experimental paradigm, which combined implicit learning and masked priming.

PhD project
“After my work with Judith, I drafted a PhD proposal for an opening at the ACLC. Initially I did not get this position, but then luckily for me someone dropped out – and I was their second choice. They asked me if I was still interested, and I accepted the job, which was really a dream come true for me. In this project I decided to return to the EEG work that I had learned about back in Chile, which I used to implement an experimental paradigm called mismatch negativity – something I had learned about in my master’s programme. Mismatch negativity is a very old EEG component, but in the 2000s it was discovered that this component was sensitive to the strength of representation of a word. Say that you hear a string of sounds like c and a, ‘ca’, in that case ‘ca’ might not correspond to any existing word. However, if you add the t you get the word cat, which does exist. The MMN component seems to be sensitive to that difference – so it can distinguish meaningless strings of sound from sounds which form words. The amplitude of the component in microvolts reflects how consolidated the word was in the mental lexicon.“

“Initially, in my PhD I wanted to find evidence for the idea that grammar is a real object of the physical world, which sounds like a very ambitious plan. I had been getting tired of all the discussions about whether grammar is just a descriptive tool, or something that exists in the real-world, and I wanted to solve this question once and for all. That is why I started out to investigate the ontological status of grammar. Usually, when we talk about grammar at the within-word level , we refer to morphemes or suffixes, and the rules that we use to combine them. Grammar at the between-word level also makes use of rules to mark agreement relation between otherwise independent words (such as the ‘-s’ in he walks.  That would be morphology and syntax. And here is when the MMN comes back: I thought, if it can distinguish whether sound strings are words or not, perhaps it is also possible for it to distinguish whether existing words are grammatically complex or not. So is the memory trace of a word the same regardless of its grammatical complexity? Or are grammatically complex forms stored and processed differently?”

“There was already some previous work on the MMN and syntactic stimuli. This work showed that if a word was a part of a larger syntactic context, the amplitude of the MMN component was reduced. The only problem with these studies (previous studies of Dutch morphological processing) was that they relied on null results to draw conclusions about the lack of grammatical processing in the Dutch lexicon. But as we know now, a null-result should not be interpreted. But more importantly, previous studies on Dutch processing systematically lacked appropriate experimental manipulations to tap into early, unconscious, automatic processing. They used lexical decision tasks without a critical manipulation called “masked priming”. Without this manipulation, the responses to lexical decision tasks can only reflect the aftermath of conscious lexical processing, but it cannot tap into  unconscious lexical processing as it occurs. Given the ample evidence that grammatical processing occurs very early after the processing of sounds, it is not surprising they never found evidence of the storage and processing of grammatical representations in Dutch. So there were a lot of methodological flaws in the older research projects, which I wanted to fix in mine.“

 “My plan was to investigate Dutch using the MMN to study the ontological status of grammar, but my results did not fit the expected patterns of MMN responses, neither in the direction of my hypothesis nor in the opposite direction. The standard interpretations of MMN effects did not seem to apply to Dutch. As I mentioned, the MMN is supposed to show the strength of word representation, a word that is often used, should be strongly entrenched and would thus yield a larger response. I replicated that effect, showing that the MMN can distinguish words from pseudowords. But then I also compared high- and low-frequency words as well, and this did not yield the expected results: I found a statistically significant reversed effect. Which makes no sense, because no theory of MMN predicts this reversed effect (Only Taft (2004) predicts a reverse frequency effect, but not like the one I got). Moreover,  although I found evidence of grammatical processing as the word-internal level (morphological processing through decomposition) I did not find a syntactic effect in the MMN with syntactic agreement relation of number, which is strange, because no theory predicts the MMN is not sensitive to syntactic regularities (the controversy lies at the morphological level). Because of this I have started to doubt whether the MMN really detects memory traces. The interpretation of the MMN as a probe of the strength of memory traces has not been reconciles (to the best of my knowledge) with the older interpretation of the MMN as a probe of acoustic change detection”

"This does not mean that my research was not fruitful. My results showed that morphological words in Dutch are stored differently than monomorphemic words, and that different mechanisms are involved, although I  don’t have all the evidence I would like to in order to make this case more strongly. I also found that the MMN varies greatly across the brain, depending on the anterior-posterior axis, and I also found some laterality effects – there is a right-hemisphere effect for monomorphemic words and a left-hemisphere effect for some syntactically complex words. And most research does not take this into account.  Nonetheless, there is still a lot of research to be done.”

After the PhD
“After my PhD, I would love to continue with this kind of research here in Amsterdam and write a post-doc proposal. I have been in contact with a researcher at the Swammerdam Institute for Life Sciences and we have discussed the idea of studying patients with refractory epilepsy, which means that they do not respond to treatment. Therefore they need to do open brain surgery, and if you receive permission you can often add another electrode for research purposes. If I got that permission, I could do MMN recordings of individual cells. And memory traces have been shown to be stored in single cells.“

“I am currently also trying to make a living as a musician, and it would be really nice if that worked out. I play the contrabass in a jazz band and the electric bass guitar in another band, but also as a session musician in the jazz sector. Amsterdam is a good place to make music, but not if you want to get paid for it. With the jazz band I do a lot of busking at Spui, which I have done throughout my PhD, but we also play at bars. My other band is louder and better suited for festivals or bigger venues. We just released a new album Viaje al Centro del Ritmo. You can listen to this on Lola’s Dice’s website.”
 

SMART Workshop – Doing Experimens in Theoretical Linguistics
“On the 4^th of April this year I will be co-organizing the SMART Workshop Doing Experiments in Theoretical Linguistics together with Jan Don. This workshop aims to inspire researchers to work on theoretically motivated experiments – something that, as I have mentioned, is not always the case. The idea for this workshop came to us at a conference that we both attended. We were not satisfied with the presentations, because a lot of researchers used empirical methods, which is good, but these methods did not always seem to be well-grounded in hypothesis testing. People know that you can do a lexical decision task, or use EEG, and get excited to use these tasks – but a lot of their work is not clearly embedded in a theoretical framework. Moreover, the statistics of their work is often seriously flawed – incorrect p-values, exploratory research which is presented as inferential research – to sum up, problems that can be avoided. And we thought – wouldn’t it be nice if there was a congress that aimed at theoretically motivated, nicely designed, empirical linguist research? We thought about this for a while, and then the opportunity for the SMART Workshop came up, which gave us the chance to finally realize our plan.”

“We think our Workshop really resonated with the linguistic community. We had to decline more than ten abstracts, which is impressive for a one-day workshop, and some big names handed in contributions. For example, our keynote speaker is Colin Phillips from the Maryland Language Science Center, whose work is perfect for our workshop. He has a lot of very well-designed and well-motivated experimental work on linguistic theory, by which I mean theory of grammar. Apart from that there is a nice mix of consolidated researchers, as well as young, new researchers, among whom not only myself, but also Ava Creemers, University of Pennsylvania and Mirjam de Jonge from the University of Amsterdam, who has done some interesting EEG work about phonological representation. So the Workshop will be a great opportunity to unite new and established people who are all working on the same things, not only for the presenters, but also for the audience. We are really looking forward to it and invite everyone to come.”
More information about the workshop can be found on the SMART Website