My emails with Marge Blanc
Part 2: My emails with Marge Blanc about her research
In part one of this series, I wrote about the reasons why I do not consider any of Marge Blanc’s publications to be research. As I stated in that post, plenty of valuable things are not research, but it is important that if we call something research, that it is actually research. It matters that what we say is true.
I regularly bring up my concerns about calling Ms. Blanc’s book and magazine articles research on various social media platforms and in talks. Marge and I have had many public interactions in the comments of various Instagram posts, and at one point last year, she asked me to email her so we could talk more about my questions about her research. On April 1, 2025, I sent her the following email:
Dear Marge,
I run the Instagram account @language_processing. I'm reaching out in response to a comment you left on one of my posts, where you invited me to email you so we can talk about the evidence that came out of your 15 years of research. I would love to see that evidence. Before you send it along, though, I want to make sure I'm in compliance with all applicable research rules. So could you please share your procedure for obtaining and documenting informed consent for research participation and any IRB approval documents you may have?
Thank you! I look forward to hearing from you.
Sincerely,
Kathleen
The following day, on April 2, 2025, she replied:
Hi Kathleen,
My intention was to actually have true dialogue, or even some kind of public presentation together, not to specifically send you "evidence" that you might use in ways that I have no idea about. And as far as sharing "evidence" from that long ago with so many people to someone who has openly critiqued what their children were a part of it's not something I would feel comfortable asking them to do.
Let's examine other ways of dialoguing.
Thanks for reaching out, however.
Marge
Her use of scare quotes around “evidence” really stood out to me. I would also think that sharing evidence with a critic is the perfect way to convince them of your point of view. I was sympathetic to her concern about sharing data without knowing what I would do with it, so that same day, I replied:
Hi Marge,
Thank you for clarifying what you were offering. I'm open to a public discussion. Would you like to find a time to chat in an Instagram live? My schedule is pretty flexible next week and again in May.
I totally understand about data sharing - as a researcher and clinician, I appreciate that you want to protect the privacy of your participants and clients. If you would consider sharing your data in the future, I'd be happy to sign a data use agreement outlining what I could and couldn't do with the data. I'll go ahead and suggest what analyses I had in mind. You're welcome to share my analysis suggestion with anyone else interested in analyzing your data. There are two things I'd be interested in doing with your 15 years of longitudinal data:
For each of the 85 clients, making a graph showing the proportion of utterances at each NLA stage over time.
For each stage, I'd like to fit a mixed effects model showing the change over time in the proportion of utterances at that stage. So for stage 1, I'd analyze how a participant's age and time in NLA therapy predict the proportion of stage 1 utterances. Then I'd do this for each of the other five stages. This kind of statistical analysis accounts for individual differences and could be used to estimate growth trajectories - I know many families have questions about how long children typically spend in a given stage, and this kind of analysis could provide some general answers to that.
In order to conduct those analyses, I'd just need summary data on each language sample - the child's age and percent of utterances in each stage. To protect the privacy of the kids, I'd ask you to create pseudonyms or numerical IDs for each child.
If you're interested in working together on this analysis, I'd be willing to agree to the following at a minimum:
Sharing my results publicly, no matter what. I've said multiple times on Instagram that I'm committed to changing my mind when presented with evidence. If the results of my analysis support what you say about NLA, I will still share the results.
Making the code (but not necessarily the data) I used to analyze the data publicly available, so anyone who wants to critique my methods can do so.
Not sharing the data with anyone else.
Sharing the results with you a few weeks before making any results public, so you would have time to check my work and prepare your own interpretation.
Not completing any analyses beyond the ones we agree on, and destroying my copies of the data after completing the agreed-upon analyses.
It sounds like you're not okay with sharing your data at this time, and I totally respect that. I just wanted to address your very valid concern about what I'd do with your data by explaining what I would do with the data.
Thanks,
Kathleen
I’m sure that the data analysts reading my email have thoughts on my proposal, and I would love to hear them. Re-reading this a year later, I would make some revisions to my second analysis idea. If the language sample data ever becomes available, I’m sure methodologists more sophisticated than I would be happy to offer suggestions for an improved analysis plan. Marge replied that same day:
Kathleen, I love your idea, with some caveats, but it would be impossible to do it that way this many years hence. Lots of reasons,,,but I have another idea.
My team and I (and other teams) are collecting data through the stages all the time. And I think it could be analyzed in the way you're describing if we really set out to save that kind of data explicitly. I would honestly love to do that! My hesitations are the time it would take for me, because I have about 50 projects going on right now, and secondly, that the kind of analysis you are talking about might well lead to parents (and others, e.g. funding sources) to look for 'results' that turn into expectations.
Could we have a zoom meeting about this? I would love to read more about your thinking, but have multiple immediate deadlines that would keep me from meeting right away. And when you're write more, could you tell me a little bit about your experience, and whether you've ever been in the company of someone who is developing language in the gestalt style? This is actually where I was going originally, because if you've not witnessed it, it's really hard to believe at a purely theoretical level.
Thanks for the thinking!
Marge
I was immediately struck by the concern that my proposed analysis might lead to parents and funding sources to look for results and expectations. Parents in the NLA Facebook group have literally asked for the kinds of research results that allow estimates of rates of progress. Why the fear of sharing that information? Marge’s interest in my clinical experience also interested me. While I understand the power of seeing language development unfold in front of you, the whole point of conducting and sharing research data is so that we don’t all need to see clients do something to believe it happened.
I also thought about her comment that we could do the analysis if she and her group “really set out to save that kind of data explicitly.” That makes me wonder how the data from her 15 years of longitudinal clinical research was saved and organized in a way that allowed Marge to calculate summary statistics as described in Addendum #3, such as “All 85 progressed from at least one stage to the next while they were in the clinic” and “The vast majority (more than 80%) progressed from Stage 1 or 2 to at least a solid Stage 4.” So on the morning of April 3, I wrote back:
Hi Marge,
I've been an SLP for 10 years, mostly in pediatrics. I've worked with kids with a huge range of diagnoses, including autistic kids who primarily communicate using delayed echolalia. I remember one little boy who had a script he picked up from Siri - "Siri is not available right now. Please try again later." He used it to help him self-regulate when he had to wait for something he wanted. Of course, we never withheld things just to make him practice waiting, but life is full of natural opportunities to wait. His mom was great and we were on the same page that his scripts were meaningful and communicative. So to answer your question about whether I've been in the company of someone developing language in a "gestalt" style, the answer is yes - I do not doubt that there are children whose expressive language trajectories begin with a lot of delayed echolalia and move towards self-generated language. I've also watched several of your webinars, both at ASHA a few years ago and on the ASHA Learning pass, and the free ones on the CDC website, so I've seen the case studies presented there as well. I do not think it's necessary to have a Zoom just to talk through what this language development looks like.
As I've shared on Instagram, what I'm questioning isn't whether some kids use echolalia or whether it's meaningful. Based on the research on language processing, I do not think that differences in processing styles are what explains echolalia use in some kids, and while language samples are a great way to conduct language evaluations and set treatment goals, they do not offer reliable insight into language processing. I also question specific aspects of NLA-based therapy that are not consistent with, or contradict, existing evidence. I'll note that following the child's lead, modeling declarative language, and honoring all forms of communication are among the evidence-based approaches that I support and use. I think I lay out my thoughts pretty clearly on Instagram (such as posts here, here, here, and here), so I won't summarize further here, but if there are specific things I've written that you wanted to chat further about I'd be happy to.
And yes, I'd be happy to talk on Zoom. One request though - I'd like to record the call so that we both have a record of what we talked about. Happy to hear what you want to chat about, and I can share thoughts on ways to organize your data collection to make future analyses easier. I do a lot of data science and what we call "data hygiene" (basically, making a plan to organize data in a way that is maximally useable) and I'd be happy to share my skills pro bono in this instance. I love that you are so generous with your time supporting families around the world in learning to honor and celebrate their child's unique communication, and if there is a way I can support in analyzing your years of data so that parents can also have access to the results of your research, I'm happy to do that.
One question: what format is your current data in? I might be able to write a computer program to automatically arrange it into a more analyzeable form, which would mean you could analyze your existing data without additional work.
Best,
Kathleen
On April 12, 2025, Ms. Blanc wrote back:
Hi Kathleen,
Yes, I do think we should have a zoom call. As far as recording for both of our personal sakes, that would be fine, as I would trust that nothing either of us say would be taken out of context and/or shared. Rereading your IG posts is interesting, but your context is so different from mine, that I feel the only fair way to represent me is to represent me. And, so far there is nothing of my context in your posts. Perhaps we may not find a way to truly connect, but I don't want to give up before we've actually tried.
This last year has produced the only critiques of NLA that I have witnessed in 20+ years, and I have wondered why the critiques of NLA have begun by looking at NLA through a completely different lens from the one that I have. So given that you are the first of the group who has been willing to have a real conversation with me, I do want to at least have one conversation in real time together to see if we can bridge perspectives — not to necessarily agree, but just to bridge perspectives.
Finding a time for a zoom call might be a bit tricky for both of us, so might I ask you what your life is like in terms of availability. I am in the central time zone, so mid day on Wednesday or Thursday is often a good time for me. How about you? I would have to plan ahead awhile, as I have regular meetings and webinars scheduled most days.
I do look forward to connecting.
Thank you,
Marge
We then had some back-and-forth about scheduling.
