Early and Late models of auditory attention Flashcards
(17 cards)
Broadbent’s Bottleneck Filter Model of Attention (Early model of attention)
1958
All auditory input enters the sensory register
There’s a “bottleneck” that only lets some information through to be processed further
The selective filter then chooses what can be processed further (through the bottleneck) based on PHYSICAL PROPERTIES such as loudness and pitch
Only the selected info reaches our higher level processing and working memory
The rest is filtered out and ignored
Treisman Attenuation Model (Late model of attention)
1964
Instead of completely blocking out unwanted information, the brain turns down the volume (attenuates) on things we’re not focusing on.
This means unattended information is still processed but more weakly.
Important or meaningful words (like your name) in the background can still get through if they’re strong enough.
So, attention works like a volume control, not an on/off switch.
Difference between early and late models of attention - processing
Early - Attention operates as a filter at an early stage of processing, before the stimulus is fully analysed for meaning
Late - All incoming stimuli are processed to some extent before selection occurs
Difference between early and late models of attention - relevant info
Early - Irrelevant info is filtered out at an early stage, and only selected information is further processed
Late - Selection of relevant info happens after the stimulus has been analysed for meaning, closer to the response stage
Difference between early and late models of attention - filtering of stimuli
Early - Stimuli are filtered based on physical characteristics (location/pitch) before semantic processing occurs
Late - Stimuli are not filtered out based on physical characteristics but rather based on their relevance to current goals or tasks
Compare and contrast early and late models of attention - Timing
Compare - Both models aim to explain how attention selects info from the environment
Constrast - The main difference lies in the timing of selection:
Early selection models propose that selection occurs at an early stage of processing
Late selection models propose that selection occurs at a later stage
Compare and contrast early and late models of attention - what to prioritise
Compare - They both acknowledge the limited capacity of attention and the need to prioritise certain stimuli over others
Contrast - Early selection models emphasise the role of physical characteristics in determining what is attended to. Whereas late selection models focus more on semantic or contextual factors
What is a breakthrough - Treisman (1964)
Top-down processes, like our expectations, play an important role. In a shadowing task (where someone repeats what they hear in one ear), people sometimes say a word from the ear they were supposed to ignore. This usually happens when that word fits well with what they’re hearing in the ear they’re focusing on.
Deutsch and Deutsch (1963)
Argued all stimuli are fully analysed, with the most important or relevant stimulus determining the response. Thus, they placed the bottleneck much later in processing than did Broadbent.
Compare Broadbent, Treisman, Deutsch & Deutsch: unattended input
Broadbent’s approach predicts little or no processing of unattended auditory messages. In contrast, Treisman’s approach suggests flexibility in
the processing of unattended messages, whereas Deutsch and Deutsch’s approach implies reasonably thorough processing of such messages.
Limitation of early and late models of attention - Importance of mechanisms underlying suppression
De-emphasised the importance of suppression or inhibition of the unattended messages: these theories focus mainly on how attention selects certain information to process, but they don’t give enough focus to the idea of actively suppressing or inhibiting distracting information. Research, like that from Horton et al. (2013) and Schwartz & David (2018), shows that the brain doesn’t just ignore distractions—it actually reduces or blocks out their signals, even in early brain areas like the primary auditory cortex.
Don’t fully explain how the brain actively blocks out distractions
Limitation of early and late models of attention - top-down processing
Classic theories often assume attention works in a fixed way (e.g., always early or always late), but newer research shows that attention is more flexible and strongly influenced by top-down factors like goals, context, and expectations.
Don’t explain how attention can adapt based on the situation or task.
Why can the early model of attention explain Cherry’s Dichotic Listening Task
We have limited capacity of auditory input that our brain can pay attanetion to
The auditory stimuli that makes it through to the bottleneck is what gets processed into higher level areas in the brain - working memory and consciousness
These two theories explain how we tune into only one auditory input, as what we aren’t interesting in doesn’t go into higher level processing
How can Triesman’s late model of attention explain the cocktail party phenomenon
Treisman concept of the attenuator explains how we can pay attention to what conversation, but our focus can shift when we hear our name etc. This shows that the other auditory input is still there, it just has had its volume turned down. Its volume can be turned up if something catches our attention
Main comparison of the early and late models of attention
Treisman’s model also explains how we can stay focused on one source of sound while ignoring others. Instead of completely blocking out the unattended information, the model suggests it’s still processed at a lower level—like the volume has been turned down or attenuated. This means we don’t fully ignore it, but it’s less likely to reach our awareness unless something important, like a familiar word, grabs our attention.
Evaluation of Treisman’s theory
Later model so does overcome some of the problems we find with Broadbent’s early model.
Explains cocktail party phenomenon really well
But there is a lack of explanation - what exactly happens at the attenuator is not explained
What is Broadbent’s model based on
Evidence from the Dichotic listening task
